experimental investigation of r134a flow in a …etd.lib.metu.edu.tr/upload/12612952/index.pdf ·...
TRANSCRIPT
EXPERIMENTAL INVESTIGATION OF R134A FLOW IN A 1.65 MM COPPER MINITUBE
BİLGEHAN TEKİN
FEBRUARY 2011
EXPERIMENTAL INVESTIGATION OF R134A FLOW IN A 1.65 MM COPPER MINITUBE
A THESIS SUBMITTED TO THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF MIDDLE EAST TECHNICAL UNIVERSITY
BY
BİLGEHAN TEKİN
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF MASTER OF SCIENCE IN
MECHANICAL ENGINEERING
FEBRUARY 2011
Approval of the thesis:
EXPERIMENTAL INVESTIGATION OF R134A FLOW IN A 1.65 MM COPPER MINITUBE
Submitted by BİLGEHAN TEKİN in partial fulfillment of the requirements for the degree of
Master of Science in Mechanical Engineering Department, Middle East Technical
University by,
Prof. Dr. Canan Özgen Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Suha Oral Head of Department, Mechanical Engineering
Assoc. Prof. Dr. Almıla Güvenç Yazıcıoğlu Supervisor, Mechanical Engineering Dept., METU
Prof. Dr. Sadık Kakaç Co-Supervisor, Mechanical Engineering Dept., TOBB ETU
Examining Committee Members:
Prof. Dr. Rüknettin Oskay Mechanical Engineering Dept., METU Assoc. Prof. Dr. Almıla Güvenç Yazıcıoğlu Mechanical Engineering Dept., METU Prof. Dr. Sadık Kakaç Mechanical Engineering Dept., TOBB ETU Assoc. Prof. Dr. Derek Baker Mechanical Engineering Dept., METU Dr. Hüsnü Kerpiççi Research & Development Expert, Arçelik
Date: 04.02.2011
iii
I hereby declare that all information in this document has been obtained and presented in accordance with academic rules and ethical conduct. I also declare that, as required by these rules and conduct, I have fully cited and referenced all material and results that are not original to this work. Name, Last name : Bilgehan Tekin Signature :
iv
ABSTRACT
EXPERIMENTAL INVESTIGATION OF R134A FLOW IN A 1.65 MM COPPER
MINITUBE
Tekin, Bilgehan
M.Sc., Department of Mechanical Engineering
Supervisor : Assoc. Prof. Dr. Almıla G. Yazıcıoğlu
Co-supervisor : Prof. Dr. Sadık Kakaç
February 2011, 154 pages
This thesis investigates the refrigerant (R-134a) flow in a minitube experimentally.
The small scale heat transfer is a relatively new research area and has been in
favor since the end of 1970’s. Refrigerant flow in mini- and microscale media is a
potential enhancement factor for refrigeration technology in the future. For the
forthcoming developments and progresses, experimental studies are invaluable
in terms of having an insight and contributing to the establishment of
infrastructure in the field in addition to leading the numerical and theoretical
approaches. The studies in the literature show that low mass flow rate and
constant wall temperature approach in minitubes and minichannels were not
among the main areas of interest. Therefore, an experimental set-up was
prepared in order to perform experiments of two-phase refrigerant flow in a 1.65
mm diameter copper minitube with the constant wall temperature approach. The
design, preparation, and modifications of the experimental set-up are explained
in this thesis. Two-phase flow and quality arrangements were done by pre-
heating the refrigerant at saturation pressure and the constant wall temperature
was achieved by a secondary cycle with water and ethylene glycol mixture as the
working fluid. The heat transfer coefficient and the pressure drop for the two-
phase flow with varying quality values and saturation temperatures of the
refrigerant were calculated and compared with the results available in literature.
Keywords: Minichannels, forced convection heat transfer, R134a, two-phase flow, experimental study.
v
ÖZ
1.65 MM’LİK BİR MİNİBORUDA R134A AKIŞININ DENEYSEL OLARAK
İNCELENMESİ
Tekin, Bilgehan
Yüksek Lisans, Makina Mühendisliği Bölümü
Tez Yöneticisi : Doç. Dr. Almıla G. Yazıcıoğlu
Ortak Tez Yöneticisi : Prof. Dr. Sadık Kakaç
Şubat 2011, 154 Sayfa
Bu tez bir miniboruda soğutkan (R-134a) akışını deneysel olarak incelemektedir.
Küçük ölçeklerdeki ısı transferi göreceli olarak yeni bir araştırma alanıdır ve
1970’lerin sonlarından bu yana revaçtadır. Mini ve mikro ölçeklerdeki soğutkan
akışı soğutma teknolojisinin geleceği için potansiyel bir iyileştirme faktörüdür.
Yaklaşan belli gelişme ve ilerlemeler için, nümerik ve teorik çalışmalara öncülük
etmesinin yanı sıra konunun iç yüzünün anlaşılması ve alandaki altyapının
kurulumuna katkıda bulunulması konularında deneysel çalışmalar çok değerlidir.
Literatürdeki çalışmalar göstermektedir ki miniboru ve minikanallardaki düşük
debi ve sabit duvar sıcaklığı yaklaşımı genel ilgi alanları arasında değildir. Bu
sebeple, 1.65 mm çaplı bakır miniborudaki iki fazlı soğutkan akışının sabit duvar
sıcaklığı yaklaşımıyla deneylerinin yapılması için bir deney düzeneği
hazırlanmıştır. Deney düzeneğinin tasarımı, hazırlığı ve yapılan değişiklikleri bu
tezde açıklanmaktadır. İki fazlı akış ve kuruluk dereceleri ayarlamaları doyma
basıncındaki soğutkanın ön ısıtıcıyla ısıtılmasıyla olmuş, sabit duvar sıcaklığı
yaklaşımı su etilen glikol karışımı kullanılan ikincil bir çevrimle sağlanmıştır.
Soğutkanın değişen kuruluk dereceleri ve doyma sıcaklıkları için iki fazlı akışta ısı
transfer katsayısı ve basınç düşmesi hesaplanmış ve literatürdeki sonuçlarla
karşılaştırılmıştır.
Anahtar kelimeler: Minikanallar, zorlanmış konveksiyonla ısı transferi, R134a, iki fazlı akış, deneysel çalışma.
vi
To My Parents
vii
ACKNOWLEDGEMENTS
The author wants to express his sincere gratitude and appreciation to his
supervisor Assoc. Prof. Dr. Almıla G. Yazıcıoğlu and co-supervisor Prof. Dr.
Sadık Kakaç for their guidance, advices, constructive criticism, encouragements
and perception throughout the research.
The author would also like to thank Dr. Hüsnü Kerpiççi, Mr. Aydın Çelik and Mr.
Yiğit Ata Ağartan for their support, suggestions and comments.
The technical assistance of Mr. Mustafa Yalçın and Mr. Fikri Çavuşoğlu are
gratefully acknowledged.
This study was supported by The Scientific and Technological Research Council
of Turkey (TÜBİTAK) Project No. 107M504, in which Arçelik provided technical
and informative support as a project partner.
viii
TABLE OF CONTENTS
ABSTRACT .........................................................................................................iv
ÖZ ....................................................................................................................... v
ACKNOWLEDGEMENTS ................................................................................... vii
TABLE OF CONTENTS ..................................................................................... viii
LIST OF TABLES ............................................................................................... xii
LIST OF FIGURES .............................................................................................xv
LIST OF SYMBOLS ......................................................................................... xviii
CHAPTERS
1. INTRODUCTION .................................................................................. 1
1.1 DOMESTIC COOLING AND EVAPORATORS ............................... 1
1.2 MICRO- AND MINICHANNEL HEAT TRANSFER .......................... 3
1.3 MOTIVATION AND OBJECTIVES ................................................ 17
2. EXPERIMENTAL SET-UP .................................................................. 19
2.1 MAIN THEORY OF THE INVESTIGATION ................................... 19
2.2 DESIGN OF THE SET-UP ............................................................ 22
2.2.1 Constant Wall Temperature................................................ 26
2.2.2 Counter Flow Heat Exchanger ........................................... 26
2.3 EQUIPMENT ................................................................................ 26
2.4 PREPARATION OF THE SET-UP ................................................ 29
2.4.1 Structural Frame ................................................................ 30
2.4.2 Data Acquisition ................................................................. 32
2.4.3 Refrigerant (R-134a) Cycle ................................................ 34
2.4.4 Water (Water + Ethylene Glycol Mixture) Cycle .................. 42
2.4.5 Test Section ....................................................................... 46
2.4.6 The Leakage and Preliminary Tests for the Initial Design ... 48
2.4.6.1 The Leakage Tests ................................................ 48
2.4.6.2 Data Acquisition Process and Preliminary Tests .... 49
2.5 MODIFICATIONS ......................................................................... 52
ix
2.5.1 Ring and Bumper Design for the Glass Tube ..................... 53
2.5.2 Leakage Tests for the Glass Tube...................................... 56
2.5.3 Copper Tube as a Replacement for Glass Tube................. 58
2.6 CALIBRATIONS............................................................................ 60
2.6.1 RTD and Thermocouple Calibrations ................................. 61
2.6.2 Transducer Calibrations ..................................................... 61
2.6.3 Heater Loss Calibrations .................................................... 63
2.6.4 Test Section Loss Calibrations ........................................... 67
3. TWO-PHASE FLOW EXPERIMENTS ................................................ 75
3.1 APPROACH .................................................................................. 75
3.2 CONDITIONS ............................................................................... 76
3.3 FLOW ORDER ............................................................................. 80
3.4 DATA ANALYSIS .......................................................................... 82
3.5 CALCULATIONS .......................................................................... 83
3.5.1 Experimental Calculations .................................................. 84
3.5.2 Calculations with Correlations in Literature ......................... 86
3.5.2.1 Güngör and Winterton Correlation .......................... 87
3.5.2.2 Chen Correlation .................................................... 88
3.5.2.3 Bertsch Correlation ................................................ 89
4. EXPERIMENTAL RESULTS AND DISCUSSION ............................... 91
4.1 HEAT TRANSFER COEFFICIENT ............................................... 92
4.1.1 Constant Quality Results of the Experiments ..................... 92
4.1.2 Constant Mass Flux Results of the Experiments ................ 95
4.1.3 Comparison with Literature ................................................ 99
4.2 PRESSURE DROP ..................................................................... 107
5. CONCLUSIONS ............................................................................... 110
5.1 SUMMARY AND CONCLUSIONS .............................................. 110
5.2 SUGGESTIONS FOR FUTURE WORK ...................................... 111
REFERENCES ................................................................................................ 112
APPENDICES ................................................................................................. 117
A. THE SPECIFICATIONS OF THE DEVICES IN THE EXPERIMENTAL
SET-UP ............................................................................................ 117
A.1 Refrigerated Circulating Bath ...................................................... 117
A.2 Digital Gear Pump System ......................................................... 118
x
A.3 Micro Flow-meter ........................................................................ 119
A.4 Glass Tube ................................................................................. 120
A.5 Compact Pressure Transducer ................................................... 120
A.6 RTD Probe ................................................................................. 121
A.7 Thermocouple Wire .................................................................... 121
A.8 Rotameter................................................................................... 122
A.9 Differential Pressure Transducer ................................................ 122
A.10 DC Power Supply ..................................................................... 123
A.11 Data Acquisition System ........................................................... 124
B. SKETCH OF THE EXPERIMENTAL SET-UP IN PRELIMINARY
DESIGN ............................................................................................ 125
C. CONNECTIONS OF EXPERIMENTAL SET-UP IN PRELIMINARY
DESIGN ............................................................................................ 126
D. THE TECHNICAL DRAWINGS OF THE PARTS IN THE TEST
SECTION .......................................................................................... 129
D.1 Technical drawing for the flange ................................................. 129
D.2 Technical drawing for the shell ................................................... 130
D.3 Technical drawings of the Ring and the Bumper......................... 131
E. THE CALIBRATION TABLES AND GRAPHS FOR MEASUREMENT
DEVICES .......................................................................................... 132
E.1 A sample calibration curve for a calibrated RTD ......................... 132
E.2 The calibration results for the CPTs ............................................ 133
E.3 The calibration curve for the CPT 1 ............................................ 134
E.4 The calibration curve for the CPT 2 ........................................... 134
E.5 The calibration results for the DPT ............................................. 135
E.6 The calibration curve for the DPT ............................................... 135
F. THE WATER ETHYLENE GLYCOL 1:1 VOLUMETRIC MIXTURE
PROPERTIES................................................................................... 136
F.1 Specific Gravity of the mixture .................................................... 136
F.2 Specific Heat of the mixture ........................................................ 137
F.3 Dynamic Viscosity of the mixture ................................................ 138
F.4 Conductivity of the mixture .......................................................... 139
G. THE REFRIGERANT (R-134A) PROPERTIES ................................. 140
G.1 Viscosity (liquid) of the refrigerant .............................................. 140
xi
G.2 Viscosity (gas) of the refrigerant ................................................. 141
G.3 Conductivity (liquid) of the refrigerant ......................................... 142
G.4 Conductivity (gas) of the refrigerant............................................ 143
G.5 Prandtl number (liquid) of the refrigerant .................................... 144
G.6 Prandtl number (gas) of the refrigerant ....................................... 145
G.7 Surface tension of the refrigerant ............................................... 146
G.8 Specific heat (liquid) of the refrigerant ........................................ 147
H. THE RESULTS FOR A TWO-PHASE EXPERIMENT ....................... 149
H.1 The Geometry of Both Sides ...................................................... 149
H.2 Test Conditions for the Experiment............................................. 150
H.3 Refrigerant Experiment States .................................................... 151
H.4 Water side properties ................................................................. 151
H.5 Refrigerant Experimental Calculations ........................................ 152
H.6 Refrigerant Calculations of the Correlations in the Literature ...... 153
H.7 Experimental Results and Results of the Correlations in the
Literature .......................................................................................... 154
xii
LIST OF TABLES
TABLES
Table 1 The states for the calibrations of pre-heater .......................................... 65
Table 2 The pre-heater calibration curve values ................................................ 66
Table 3 Water side geometrical constraints ....................................................... 69
Table 4 Viscous heating experimental values .................................................... 70
Table 5 The water side geometry for heat loss calculations ............................... 73
Table 6 The results for the maximum heat loss to the surroundings................... 74
Table 7 The refrigerant conditions for the test section in two-phase flow ........... 77
Table 8 The average refrigerant parameters for the two-phase experiments ..... 78
Table 9 The water side parameters range for two-phase flow experiments ........ 78
Table 10 The pre-heater and the test section heat transfer rates for two-phase
experiments ....................................................................................................... 79
Table 11 Flow order part 1: from computer start-up to refrigerant pump start-up ...
.......................................................................................................................... 80
Table 12 Flow order part 2: from refrigerant pump start-up until the end ............ 81
Table 13 The refrigerant side geometry in the test section ................................. 83
Table 14 Heat transfer coefficient dependence on saturation pressure for
constant refrigerant quality ................................................................................. 93
Table 15 Heat transfer coefficient dependence on mass flux for constant
refrigerant quality ............................................................................................... 94
Table 16 Heat transfer coefficient dependence on saturation pressure for
constant refrigerant mass flux ............................................................................ 96
Table 17 Heat transfer coefficient dependence on quality for constant refrigerant
mass flux ........................................................................................................... 98
Table 18 Comparison of experimental results for the heat transfer coefficient for
constant quality and varying mass flux ............................................................... 99
Table 19 Comparison of experimental results for the heat transfer coefficient for
varying quality .................................................................................................. 102
Table 20 Comparison of experimental results with Tibiriçá [28] in terms of quality
dependence ..................................................................................................... 103
xiii
Table 21 Comparison of experimental results with Shiferaw [23] in terms of
quality dependence .......................................................................................... 105
Table 22 Dependence of pressure drop on mass flux for constant quality ....... 107
Table 23 Dependence of pressure drop on quality for constant refrigerant mass
flux ................................................................................................................... 109
Table C.1: The necessary pipes for the set-up connections ............................. 126
Table C.2: Elbows and T-connections .............................................................. 126
Table C.3: Refrigerant (R-134a) cycle connectors and pipes ........................... 127
Table C.4: Water cycle connectors and pipes .................................................. 128
Table E.1: The calibration results for the compact pressure transducers ......... 133
Table E.2: The calibration results for the differential pressure transducer ........ 135
Table F.1: Specific gravity dependence on temperature for the water – ethylene
glycol mixture ................................................................................................... 136
Table F.2: Specific heat dependence on temperature for the water – ethylene
glycol mixture ................................................................................................... 137
Table F.3: Dynamic viscosity dependence on temperature for the water –
ethylene glycol mixture .................................................................................... 138
Table F.4: Conductivity dependence on temperature for the water – ethylene
glycol mixture ................................................................................................... 139
Table G.1: Viscosity (liquid) dependence on saturation pressure for R-134a ... 140
Table G.2: Viscosity (vapor) dependence on saturation pressure for R-134a ... 141
Table G.3: Conductivity (liquid) dependence on saturation pressure for R-134a ....
........................................................................................................................ 142
Table G.4: Conductivity (vapor) dependence on saturation pressure for R-134a ...
........................................................................................................................ 143
Table G.5: Pr (liquid) dependence on saturation pressure for R-134a .............. 144
Table G.6: Pr (vapor) dependence on saturation pressure for R-134a ............. 145
Table G.7: Surface tension dependence on saturation pressure for R-134a .... 146
Table G.8: Specific heat (liquid) dependence on saturation pressure for R-134a ...
........................................................................................................................ 147
Table H.1: The geometries of both sides ......................................................... 149
Table H.2: Test conditions for the experiment .................................................. 150
Table H.3: Refrigerant experiment states ......................................................... 151
Table H.4: Water side properties ..................................................................... 151
xiv
Table H.5: Refrigerant experimental calculations ............................................. 152
Table H.6: Calculations of correlations in the literature .................................... 153
Table H.7: Experimental results and the results of the correlations in literature
........................................................................................................................ 154
xv
LIST OF FIGURES
FIGURES
Figure 1 Conventional Mechanical Refrigeration System ..................................... 2
Figure 2 Schematic view of the set-up at the first design ................................... 24
Figure 3 Schematic view of the set-up at preliminary design with states ............ 29
Figure 4 Schematic view of the set-up at preparation stage ............................... 30
Figure 5 Technical drawing of structural frame .................................................. 32
Figure 6 The photos of data logger .................................................................... 32
Figure 7 The photo of thermocouple plug panels ............................................... 34
Figure 8 The refrigerant bath and the coil heat exchanger ................................. 36
Figure 9 The refrigerant cycle (bath section) ...................................................... 37
Figure 10 The refrigerant cycle (the pump and the flowmeter before by-pass line
added) ............................................................................................................... 38
Figure 11 The photo of flowmeter and by-pass line ............................................ 39
Figure 12 The pre-heater before and after winding ............................................ 40
Figure 13 The control box and the power transducer of the pre-heater .............. 41
Figure 14 The water cycle .................................................................................. 45
Figure 15 The photo of the test section at the initial preparation ........................ 47
Figure 16 The photo of the set-up after the preliminary tests ............................. 52
Figure 17 The schematic view of the set-up in the actual experiments ............... 53
Figure 18 The adapter parts for glass - copper transition ................................... 54
Figure 19 The adapter parts mounted on the cycle ............................................ 55
Figure 20 The test section after the shell was closed with bolts and silicone ...... 56
Figure 21 The exit of the test section with solidified glue at the adapter ............. 57
Figure 22 The test section with copper tube ....................................................... 58
Figure 23 Test section with copper tube after insulation..................................... 59
Figure 24 The pre-heater section of refrigerant cycle as control volume ............ 64
Figure 25 The calibration curve of the pre-heater............................................... 67
Figure 26 Viscous heating versus Reynolds number ......................................... 71
Figure 27 Heat transfer coefficient versus saturation pressure for constant
refrigerant quality ............................................................................................... 93
xvi
Figure 28 Heat transfer coefficient versus mass flux for constant refrigerant
quality ................................................................................................................ 95
Figure 29 Heat transfer coefficient versus saturation pressure for constant
refrigerant mass flux .......................................................................................... 96
Figure 30 Heat transfer coefficient versus quality for constant refrigerant mass
flux ..................................................................................................................... 98
Figure 31 Experimental results and results of correlations in the literature for heat
transfer coefficient versus mass flux, x = 18% ................................................. 100
Figure 32 Heat transfer coefficient ratio versus mass flux, x = 18% ................. 101
Figure 33 Experimental results and results of correlations in the literature for the
heat transfer coefficient versus quality ............................................................. 102
Figure 34 Heat transfer coefficient versus quality for current experiments and
results of Tibiriçá [28] ....................................................................................... 104
Figure 35 Heat transfer coefficient versus quality for the current experiments and
those of Shiferaw [23], G = 200 kg/m2s ............................................................ 106
Figure 36 Pressure drop versus mass flux for constant refrigerant quality ....... 108
Figure 37 Pressure drop versus quality for the constant refrigerant mass flux, G
= 200 kg/m2s .................................................................................................... 109
Figure B.1 Sketch of the experimental set-up in the preliminary design ........... 125
Figure D.1 Technical drawing of the flange ...................................................... 129
Figure D.2 Technical drawing of the shell ........................................................ 130
Figure D.3 Technical drawings of the ring and the bumper .............................. 131
Figure E.1 A sample calibration curve for an RTD ........................................... 132
Figure E.2 The calibration curve for the compact pressure transducer 1 .......... 134
Figure E.3 The calibration curve for the compact pressure transducer 2 .......... 134
Figure E.4 The calibration curve for the differential pressure transducer .......... 135
Figure F.1 Specific gravity versus temperature for the water – ethylene glycol
mixture ............................................................................................................. 136
Figure F.2 Specific heat versus temperature for the water – ethylene glycol
mixture ............................................................................................................. 137
Figure F.3 Dynamic viscosity versus temperature for the water – ethylene glycol
mixture ............................................................................................................. 138
Figure F.4 Conductivity versus temperature for the water – ethylene glycol
mixture ............................................................................................................. 139
xvii
Figure G.1: Viscosity (liquid) versus saturation pressure for R-134a ................ 141
Figure G.2: Viscosity (vapor) versus saturation pressure for R-134a ............... 142
Figure G.3: Conductivity (liquid) versus saturation pressure for R-134a ........... 143
Figure G.4: Conductivity (vapor) versus saturation pressure for R-134a .......... 144
Figure G.5: Pr (liquid) versus saturation pressure for R-134a .......................... 145
Figure G.6: Pr (vapor) versus saturation pressure for R-134a .......................... 146
Figure G.7: Surface tension versus saturation pressure for R-134a ................. 147
Figure G.8: Specific heat (liquid) versus saturation pressure for R-134a .......... 148
xviii
LIST OF SYMBOLS
A ............................................................ inner surface area of the copper tube, m2
Al ............................................................................................................ aluminum
Bo ................................................................................................... boiling number
Br ............................................................................................... Brinkman number
Co .......................................................................................... confinement number
cp ........................................................................................... specific heat, kJ/kgK
Dh ........................................................................................ hydraulic diameter, m
E ............................................................................................. enhancement factor
F ................................................................................................... correction factor
Fo ................................................................................ boiling enhancement factor
G ................................................................................................ mass flux, kg/m2s
g .................................................................... gravitational acceleration, 9.81 m2/s
ℎ ........................... average convective heat transfer coefficient of R134a, W/m2K
h .....................................................................................................enthalpy, kJ/kg
I ........................................................................................................... current, mA
i .................................................................................................. latent heat, kJ/kg
k ................................................................................. thermal conductivity, W/mK
L ................................................................................ Length of the test section, m
M ...................................................................................moleculer weight, kg/kmol
Nu ................................................................................................. Nusselt number
............................................................................................ mass flow rate, kg/s
p ........................................................................................................ pressure, Pa
Pr ................................................................................................... Prandtl number
............................................................................................ heat transfer rate, W
q” ................................................................................................... heat flux, W/m2
r ............................................................................................................... radius, m
Re .............................................................................................. Reynolds Number
S ............................................................................................... suppression factor
T ..................................................................................................... temperature, K
t ......................................................................................................... thickness, m
xix
U .............................................................. Overall heat transfer coefficient, W/m2K
..................................................................................... volumetric flow rate, m3/s
Xtt ............................................................................. Lockhart-Martinelli parameter
x .................................................................................................................. quality
Greek letters
∆ ............................................................................................................. difference
µ ................................................................................................. dynamic viscosity
ρ ................................................................................................................. density
σ .................................................................................................... surface tension
θb .................................................................................................... wall superheat
Subscripts
c ................................................................................................................. copper
cb .............................................................................................. convective boiling
conv ...................................................................................................... convection
H ............................................................................................................ pre-heater
he ................................................................................................... pre-heater exit
hi ................................................................................................... pre-heater inlet
i ............................................................................................................. tube inner
ins ........................................................................................................... insulation
l ..................................................................................................................... liquid
LM .............................................................................................. logarithmic mean
nb .................................................................................................. nucleate boiling
o ............................................................................................................ tube outer
p .......................................................................................................... pool boiling
R ............................................................................................................ refrigerant
r ................................................................................................................ reduced
re .............................................................................................. refrigerant test exit
ri .............................................................................................. refrigerant test inlet
s ............................................................................................................. saturation
si ............................................................................................................ shell inner
so .......................................................................................................... shell outer
TP .......................................................................................................... two-phase
xx
V .................................................................................................... viscous heating
w ................................................................................................... test section wall
W .................................................................................................................. water
W,loss .............................................................. water side loss to the surroundings
we .................................................................................................... water test exit
wi .................................................................................................... water test inlet
1
CHAPTER 1
INTRODUCTION
1.1 DOMESTIC COOLING AND EVAPORATORS
In the present era, the demand and the need for energy are soaring with the
increase of civilization. Therefore, careful use of energy resources and increase
in the efficiency of energy transfer are worth and needed to be considered. In the
energy field, heat transfer is one of the main areas. There are several fields
where heat transfer takes place and/or is used with different aspects and subjects
in daily life, in industrial applications to name a few.
Among the areas of heat transfer, cooling is a significant concept for a wide
range of applications. For almost all fields of industrial applications and domestic
appliances, it has a great importance either directly or indirectly. For instance; in
refrigerators, cooling is the main purpose to protect the food and beverages
inside, while in automobiles, for better performance of engine and for the engine
parts not to be worn or damaged, cooling has an indirect but important effect.
When the domestic cooling is taken into consideration, refrigerators and air
conditioners are the main devices used. They differ in terms of the aim,
mechanism of the cooling and the energy consumption rate. Refrigerators
compose more than 13% of total electricity consumption in households [1], which
is an important factor for the thesis in terms of the motivation; and will be
explained in the forthcoming parts. In this thesis, evaporator of a refrigerator,
specifically, alternatives for the tubing within, is the subject of interest.
2
For a household refrigerator, the cooling process takes place in a refrigeration
cycle. Figure 1 shows conventional vapor compression refrigeration cycle [2]. In
the evaporator part, heat is absorbed from refrigerator food compartment; and the
low pressure, low temperature liquid refrigerant is evaporated at a relatively
constant pressure and temperature.
Figure 1: Conventional Mechanical Refrigeration System
In the conventional refrigerators, the evaporator is composed of circular tubes
usually made from copper and these tubes are conventional (macroscale)
channels [3]. In this thesis using minitubes instead of conventional channels in
the evaporator is investigated experimentally by designing and building a new
experimental set-up with the constant wall temperature approach for the two-
phase flow of refrigerant-134a (R-134a) in a minitube (circular) with a diameter of
1.65 mm.
Expansion valve
Heat rejected to
air or water
Condenser
Evaporator
Refrigeration Load
Motor
High pressure,
temperature liquid
High pressure,
temperature gas
Low pressure,
temperature liquid Low pressure,
temperature gas
Compressor
3
1.2 MICRO- AND MINICHANNEL HEAT TRANSFER
Heat transfer in small scales is one of the relatively new fields that has been
attracting researchers’ interest. There are different approaches for the naming of
small scale tubes and channels. According to Kandlikar [3]; when tubes or
channels have hydraulic diameter (Dh) larger than 3 mm, they are called
conventional channels, while minichannels have 3 mm > Dh > 200 µm, and
microchannels have 200 µm > Dh > 10 µm. This thesis is based on Kandlikar’s
categorization, but in some studies, as well as the industry, the ranges given
above were not considered the same by the authors and they called all small
channels microchannels. In order not to change the original terminology; when
studies from literature are presented, the authors’ naming for the channels is
used in this text.
For the last three decades, micro- and mini-scale heat transfer studies have been
increasingly popular. One of the very first studies was conducted by Tuckerman
and Pease [4]. They experimentally studied microchannel water flow for cooling
the CPU of a computer and showed that heat transfer increases in
microchannels, compared to macrochannels. This study attracted the interest of
many researchers to examine the heat transfer at mini- and micro-scales.
Moreover, the fact that the correlations for the conventional channels were not
valid for the mini- and micro-channels led to an instant necessity for new studies
about this subject.
As mentioned in the Introduction, and as it is known from basic principles of
thermodynamics, there is two-phase flow in the evaporators. In the refrigerators,
refrigerants are used because they have lower boiling temperatures and high
heat capacities. The facts that heat transfer and heat transfer coefficient are
much larger for two-phase flow than for single-phase flow and that small scale
heat transfer is greater than large scale (conventional) heat transfer directed the
studies to be focused on refrigerant flow boiling in mini- and microchannels.
Heat transfer for a flowing fluid in an internal flow is characterized by several
variables. These are mainly the bulk fluid temperature and pressure, mass flow
4
rate / fluid velocity (Reynolds number), geometry (channel height, width and
length / tube diameter and length), channel port number (single- / multi-channel),
flow type (laminar / turbulent), flow phase (single- / two-phase), boundary
conditions (constant wall temperature / constant heat flux), pressure drop (friction
factor), heat transfer coefficient (Nusselt number).
Several researches about the effects of the above mentioned variables on the
heat transfer for minichannels, flow boiling of refrigerants, and both have been
performed. The previous studies and their results directed the idea of the current
experimental set-up and presented the necessity of this field. Among hundreds of
invaluable scientific studies, some were examined and they are mentioned in the
forthcoming section in the relation about the present thesis.
For a single-phase internal flow, the flow regime is an important parameter for
heat transfer performance. There are two main flow regimes, designated by the
Reynolds number (Re), which are laminar and turbulent flow. Reynolds number is
a dimensionless quantity, which may be thought of as the ratio of the inertial
forces in the fluid to the viscous forces. For internal flow in conventional
channels, flows with Re < 2300 are designated as laminar. Although many text
books inform that when Re > 2300, the flow is called turbulent; there is actually
no sharp transition. Instead, there is a flow regime which is neither laminar nor
turbulent, and it is called transition flow. Choi et al. [5] made a research on dry
nitrogen flow in microchannels, about the effect of flow regime on heat transfer in
microchannels. The authors compared the flow regime effect on heat transfer for
long micro-tubes with diameter between 3 – 81 µm and proposed a correlation for
Nusselt number (Nu) as a function of Re for laminar flow. They concluded that
laminar flow had better performance in heat transfer than turbulent flow.
Moreover, Harms et al. [6] studied the thermal and hydraulic properties of the
flow of deionized water for two micro-channel configurations (single- and multi-
channels) theoretically and experimentally and found the critical Reynolds
number as 1500. They also concluded that the laminar flow was better in terms of
heat transfer performance than turbulent flow similar to Choi et al. [5].
5
Flow geometry also has an important effect on the internal flow in mini- and
microchannels. The tubes are mainly considered as circular and non-circular, in
terms of the cross sectional shape. Non-circular tubes are traditionally called
channels in text books. These channels can have various shapes; triangular,
rectangular, square, wide parallel channels with small gaps in between, to name
a few. One other parameter to be considered about the flow geometry is the
number of ports in which fluid flows. When there is one tube / channel, it is called
a single channel; on the other hand, the flow can be divided into several ports
(multi-port microchannels) to increase the wetted perimeter and decrease the
hydraulic diameter. Since the beginning of the research of internal flow in mini-
and microchannels, several studies have been done about the effect of flow
geometry on the heat transfer performance and pressure drop. Jung and Kwak
[7] observed water flow in 5 parallel rectangular microchannels with 3 different
hydraulic diameters. The authors kept the channel length constant and evaluated
the friction factor. They observed that the values were close to the theoretical
values for macroscopic correlations, whereas the Nusselt number relation
deviated and was strongly dependent on Reynolds number. Therefore, the
authors offered a new correlation instead of using constant value of Nusselt
number for laminar developed flow as in macroscale tubes and channels.
Moreover it was shown that the heat transfer coefficient was linearly dependent
on the wall temperature and fluid properties; such as viscosity, which affects the
heat transfer mechanism significantly. Park and Punch [8] concentrated on
single-phase flow in microchannels experimentally. The aim was to investigate
the friction factor and heat transfer in arrays of microchannels for different
hydraulic diameters ranging from 106 µm to 307 µm for low Reynolds number
(69<Re<800) values. The friction factor results matched the correlations for
conventional channels whereas the Nusselt number values deviated significantly.
As a result, the authors proposed a correlation including Nusselt, Reynolds,
Prandtl (Pr), and Brinkman (Br) numbers. Brinkman number is a dimensionless
number that relates the heat conduction from a wall to a flowing viscous fluid.
Besides the flow regime and flow geometry, many studies on the single- and two-
phase flow have been performed. Heat transfer in two-phase flow is known to be
6
superior compared to single-phase flow. Studies were focused on the effect of
different parameters on two-phase mini- and microchannel flow.
Kandlikar [9] reviewed the previous studies on mini- and microchannel heat
transfer and pressure drop. The comparison of the mentioned channels with the
conventional-sized channels/tubes was handled and flow boiling characteristics
were considered. Also, the author pointed to the fact that the configuration – one
or multi ports – had an important effect on heat transfer and pressure drop. His
study focused on relevant geometries, fluids, and heat/mass fluxes. He
concluded that for multi-port channels, large pressure drop fluctuations occurred
and also stated that heat transfer correlations for mini- and microchannels could
be better predicted from the macroscale correlations as a first estimate.
Moreover, he mentioned that the deviations in heat transfer and pressure drop of
single and multichannels were caused by the fluctuations and flow instabilities.
Although this article is relatively older (2002) than the others, it is beneficial to
have an insight on the differences between single- and multiport channels and
micro- and macroscale flows. In addition, an organized literature review is
beneficial when comparing the current study with those in the literature.
Qu and Mudawar [10] investigated the boiling heat transfer of water flowing in
parallel microchannels. The cross-section of each channel was 231 µm × 713 µm
and there were 21 channels in total. For different inlet and boundary conditions,
they performed experiments and concluded that contrary to macroscale flow, the
heat transfer tended to decrease with increasing equilibrium quality. Several
available correlations from literature were analyzed and it was found that none of
them succeeded in predicting the experimental results, since most of their ranges
of operation did not include the experimental conditions.
Saitoh et al. [11] performed experiments with three small diameter tubes (0.51,
1.12, and 3.1 mm inner diameter) using R-134a as the working fluid, which is also
used in the current research. The authors examined the local heat transfer
coefficient and pressure drop for all three tubes for different heat and mass
fluxes, evaporating temperatures, and inlet qualities. The main goal for the
experiments was to investigate the effect of tube diameter on heat transfer for
7
two-phase flow conditions. They concluded from the experiment that as the
diameter gets smaller, the forced convective evaporation reduces, which means
that the experimental heat transfer coefficient increases with increasing heat flux,
but was not affected with the increasing mass flux. In addition, a dry-out in the
lower quality region has been observed as the tube diameter decreased. On the
other hand, after calculating the frictional pressure drop with the homogeneous
model and the Lockhart-Martinelli correlation, and comparing with the
experimental results, the authors realized that as diameter decreases, one can
better predict the frictional pressure drop by using the homogeneous model
rather than by the Lockhart-Martinelli correlation.
Bertsch et al. [12] worked with refrigerants and performed experiments in copper
microchannel cold plate evaporators. In addition to R-134a, the authors used R-
245fa as the working fluid. For two different arrays of microchannels with
hydraulic diameters of 1.09 and 0.54 mm and with an aspect ratio of 2.5 for both,
they measured heat transfer coefficient for different quality, saturation
temperature, mass flux, and heat flux values. After their experiments, the authors
concluded that the heat transfer coefficient was highly dependent on heat flux
and quality while it was not affected much by saturation temperature and mass
flux in the range of interest. Also, they concluded that in the flow boiling regime,
R-134a was superior in heat transfer than R-245fa. Moreover, as the mass flux
increases, the dominance of nucleate boiling increases for higher heat fluxes.
The dependence of heat transfer coefficient on quality was also examined and
the outcome was that the heat transfer coefficient strongly depends on the quality
and there was a rapid decrease in the coefficient for qualities above 0.5. The
highest value of heat transfer coefficient was seen for the quality values between
0.1 and 0.5 depending on the geometry and flow conditions. On the other hand, it
has been reported that the smaller the hydraulic diameter, the higher the heat
transfer coefficient, however small the difference. The authors also compared
their results with several correlations in the literature and found that only a few
are applicable for these microchannels and with errors as well.
Revellin and Thome [13] studied the flow-boiling regimes experimentally and they
used an optical measurement method. Their test tubes were glass tubes, 0.5 and
8
0.8 mm in diameter. The authors carried out the experiments with R-134a and R-
245fa at different saturation temperatures and pressures. The experiments
showed four different flow patterns with their transitions. The interesting result
was that for R-134a, neither macroscale correlations nor the microscale
correlations for air/water flows were close to the results. Inlet sub-cooled
temperatures and the saturation pressures did not significantly affect the
performance. Also the different channels’ results were not much different from
each other except the flow patterns. The main difference recorded was that for R-
134a, flow regime transitions were more affected by the mass flux than for R-
245fa, which indicates that mass flux is an important issue to handle in two-phase
flow in mini- and microchannels.
Bertsch et al. [14] also worked with refrigerant R-134a as the working fluid with
similar conditions and the same aspect ratio in their previous work [12]. The
authors examined a microchannel evaporator with 17 parallel channels of
hydraulic diameter 1.089 mm (this time they did not consider any other channels
with different hydraulic diameter) and found that the heat transfer coefficient
reaches the largest value for the quality of 0.2 for all the experiments. Also, they
realized that the heat transfer coefficient decreased rapidly after the maximum
value with the increasing quality. In these experiments, one new outcome was
that the effect of saturation pressure on the heat transfer coefficient was
negligible. Moreover, it was reported that different from other microchannel
studies, the heat transfer coefficient increased dramatically not only with the
increasing heat flux but also with the increasing mass flux.
Yan and Lin [15] investigated evaporation heat transfer coefficient and pressure
drop in a small pipe. These authors also used R-134a as the working fluid. The
diameter of the pipe was 2 mm, which is close to the test pipe diameter in the
current study, and is valuable to understand how the results deviate from the
larger pipe heat transfer conditions. The authors found that for the small pipe
mentioned above, the evaporation heat transfer coefficient is about 30-80%
higher with the increasing heat flux and mass flux than that for conventional
pipes; however the pressure drop was also larger. They also offered correlations
for the heat transfer coefficient and friction factor.
9
Liu et al. [16] performed a numerical study of heat transfer in microchannels.
They were interested in 2-D modeling of the water flow in a 100 µm hydraulic
diameter microchannel for low Reynolds number flows and localized heat flux
conditions. They focused on the viscosity and thermal conductivity variation and
their effects on the heat transfer. It was revealed that the heat transfer increased
with the viscosity variation with temperature, while the axial conduction resulting
from the thermal conductivity variation seemed to be negligible except for very
low Reynolds number cases. In their study, the significant outcomes were that
the velocity field and the temperature distribution were coupled, the velocity field
was distorted with the property variations, and the cross-flow introduced in the
numerical approach affected the convection in a non-negligible manner.
Liu and Garimella [17] made experiments with water flowing inside
microchannels. There were two different channels with dimensions of 275 x 636
and 406 x 1063 µm2. The authors tried different water inlet temperatures with a
variety of mass fluxes and different heat fluxes with a maximum value of 129
W/cm2. They also tried two-phase flow but only with a maximum exit quality of
0.2. After the experiments, they calculated the convective heat transfer coefficient
and compared the values with the results of correlations for macrochannels. As in
the previous studies, the outcome is that the correlations worked for the
subcooled region, but not for the saturated flow boiling regions. Then, the authors
offered an analytical solution for the flow boiling region, which matched well with
the experiments.
Agostini et al. [18] studied the flow boiling of R-134a in minichannels
experimentally. The authors used 11 parallel rectangular multichannel straight
aluminum drawn tubes with a hydraulic diameter of 2.01 mm. The working
conditions were in the ranges of: 90 – 295 kg/m2s for mass flux and 6.0 – 31.6
kW/m2 for heat flux. The saturation pressures for the experiments were 405 and
608 kPa. The cooling at the inlet was varied from 1 K to 17 K. They concluded
that although dry-out occurred for low quality values, heat transfer for the
minichannels was greater than the heat transfer for conventional channels in the
literature.
10
Djordjevic et al. [19] examined the evaporation heat transfer of ammonia and R-
134a experimentally in a gasketed plate heat exchanger with Chevron-type
plates. Saturation temperatures were between 268 – 283 K for ammonia and 265
– 283 K for R-134a. The second fluid used in the heat exchanger was a water –
ethylene glycol mixture, which is also the secondary fluid used in the current
experiments. In the heat exchanger, parallel flow resulted in a better performance
than counter flow, contrary to the expectations of the authors. Comparisons with
the literature gave a good match in terms of heat transfer coefficient even if the
results for flow arrangements are against, and some correlations from the
literature were adapted to the heat exchanger used in the experiments.
De Rossi et al. [20] carried out flow boiling heat transfer and pressure drop
experiments for R-134a with a saturation temperature between -8.8 and 19.9ºC.
The tube they selected was 6 mm in diameter and stainless steel. Their working
mass flux and heat flux values were variable. After the experiments, they were
able to comment on how these variables affected the heat transfer coefficient and
pressure drop. They concluded that the heat transfer coefficient increased with
the increasing quality, heat flux, and mass flux, but the saturation pressure
affected the heat transfer coefficient only for low qualities. The pressure drop also
increased with the increasing mass flux and decreasing saturation pressure.
Ong and Thome [21] performed experiments on flow boiling inside a 1.030 mm
diameter channel. They chose three different refrigerants; R-134a, R-236fa, and
R-245fa. They collected data for flow boiling heat transfer and flow pattern
visualization. Similar to the other researchers, they conducted experiments with
different heat and mass fluxes and concluded that for the saturated flow boiling
heat transfer, the local heat transfer coefficient showed strong dependence on
heat and mass flux but no significant dependence on the subcooled residuals.
The authors mentioned that the changes due to these dependencies in heat
transfer matched well with the flow transitions and added that the forced
convection and nucleate boiling in microchannels was still unclear and should be
studied further.
11
Choi et al. [22] used 2000 mm long, 1.5 and 3.0 mm diameter horizontal stainless
steel minitubes in their study. They examined the flow boiling heat transfer of R-
22, R-134a, and CO2 in minichannels flow. They evaluated the heat transfer
coefficient for laminar flow with 10 – 40 kW/m2 heat flux, 200 – 600 kg/m2s mass
flux, and 0 – 1.0 quality variations. They varied heat flux values for different
experiments and investigated the constant heat flux performance. For different
fluids, the results for average heat transfer coefficient ratios, which are the ratios
of the average heat transfer coefficients of the fluid to that of R-22, were 1.0, 0.8,
and 2.0 for R-22, R-134a, and CO2, respectively. In conclusion, they developed a
new correlation for heat transfer coefficient in flow boiling heat transfer in
minitubes.
Shiferaw et al. [23] compared the flow boiling heat transfer results of flow in 1.1
mm inner diameter stainless steel tube with the results of three region flow
model. The fluid used in the experiments was R134a. The parameters changed
were mass flux (100 – 600 kg/m2s), heat flux (16 – 150 kW/m2) and saturation
pressure (6 – 12 bars). According to the results of the experiments; at quality
values below 50% and low heat and mass fluxes, heat transfer coefficient
changes with the heat flux and saturation pressure, but not with the quality,
contrary to many previous studies. Mass flux effect seemed to be negligible on
heat transfer coefficient. For the high heat flux and above 50% quality values,
heat transfer coefficient does not change with heat flux and decreases with the
increasing quality value, which was considered to be caused by dry-out. At low
pressure values, the three region model had a good match with the experimental
results, whereas it was not valid for the dry-out regions.
Matkovic et al. [24] took the condensation of R134a and R32 in 0.96 mm
diameter stainless steel minichannel into consideration and calculated the local
heat transfer coefficient by measuring the local heat flux and saturation
temperature and pressure. In the experiments, the mass flux was varied from 100
to 1200 kg/m2s and the saturation temperature was 40°C. The experimental
results came out to be consistent with the results from the condensation
equations derived for conventional channels. It has been concluded that for the
minichannel condensers, on condition that there is relatively high velocity flow
12
inside circular cross-sectional minichannel, the correlations for the conventional
channels are applicable.
Revellin et al. [25] studied the critical heat flux, which is an important subject in
boiling, in mini- and microchannels. They compared the methods used in the
literature to find the best match to the critical heat flux in small scale flow. The
database they established consisted of the experimental investigation on 9
different fluids including R134a. They proposed the best correlations and
methods for water and non-water based fluids among the ones they compared.
In et al. [26] observed the flow boiling of R134a and R123 experimentally in
circular stainless steel microtubes with an inner diameter of 0.19 mm. They
varied some parameters in the experiments, which were heat flux, mass flux,
quality, and saturation pressure. Heat flux values were 10, 15 and 20 kW/m2,
mass fluxes were 314, 392 and 470 kg/m2s, saturation pressures were 158 and
208 kPa for R123, 900 and 1100 kPa for R134a, and quality was between 0.2
and 0.85. For the two-phase heat transfer in R123 flow, nucleate boiling ran out
at low quality values and the evaporation of liquid layers around augmented
bubbles were dominant. On the other hand, for the two-phase heat transfer in
R134a flow, up to larger quality values, nucleate boiling was dominant in flow
boiling and for large quality values two-phase forced convection heat transfer was
dominant. The results for R134a were similar to the macrochannel characteristics
except the nucleate boiling development and small forced convection area. At the
low and medium (about 0.5) quality values, heat transfer coefficient highly
depended on heat flux and saturation pressure whereas mass flux and quality did
not have an important effect on it. On the other hand, at high quality values, while
heat flux and saturation pressure effects on heat transfer coefficient were
negligible, mass flux had significant effects.
Park et al. [27] acquired new experimental results for critical heat flux values for
the flow of three refrigerants (R134a, R236fa, R245fa) in two different copper
multi-port microchannels. One of the test sections was composed of 20 parallel
rectangular channels, 467 µm wide and 4052 µm deep. The other test section
was composed of 29 channels, 199 µm wide and 756 µm deep. Microchannels
13
were 30 mm long and they were heated up with electrical resistances along a 20
mm length at the mid-section. For the varying mass flux values from 100 kg/m2s
to 4000 kg/m2s, critical heat fluxes were measured to be from 37 W/cm2 to 342
W/cm2. Critical heat flux increased with increasing mass flux, and at high
velocities (high mass fluxes) the increase was observed to be less. In the 4052
µm deep channels, the critical heat flux partly increased with the inlet cooling, but
as the cross-sectional area of inlet cooling decreased; it was observed that the
effect on the critical heat flux also decreased. Depending on the flow condition
and the channel dimensions, the critical heat flux was inversely proportional to
the increasing saturation temperature. When the experimental results were
compared with the correlations, there was a good match with the ones for circular
tubes using effective diameter and mass velocity.
Tibiriça et al. [28] researched the flow boiling heat transfer of R134a and R245fa
in a 2.3 mm inner diameter stainless steel tube experimentally and also displayed
the characterization of flow model. They gathered the experimental data for the
flows in the ranges of 50 – 700 kg/m2s mass flux, 5 – 55 kW/m2 heat flux and
0.05 – 0.99 quality values. The calculated heat transfer coefficients were between
1 – 14 kW/m2K. They concluded that heat transfer coefficient strongly depends
on the heat flux, mass flux and quality.
Raja et al. [29] tested the flow boiling of a mixture composed of three refrigerants
– R-134a, R290 and R600a with mixing ratios of 91%, 4.068%, and 4.932% (in
terms of mass), respectively – in two different, 9.52 and 12.7 mm diameter,
horizontal tubes. They applied heat flux values of 2 – 8 kW/m2 to find the
behavior of heat transfer coefficient under constant heat flux approach. The inlet
temperature of the refrigerant was between -9 – 5°C and mass flow rate was in
the range of 3 – 5 g/s. The heat transfer coefficient values came out to vary in the
range from 500 to 2200 W/m2K in two tubes. They compared the results for both
tubes and found that the heat transfer coefficient in the 9.52 mm tube is 10 % –
45 % (for different flow rates) larger than the 12.7 mm tube. Moreover, they made
a comparison of heat transfer performance for refrigerant mixture and pure
R134a. When the pressure was constant along the flow, the mixture gave better
performance throughout the boiling, whereas for constant inlet temperature, it
14
was larger only for high qualities. In the experiments, the authors also discussed
the boiling mechanisms in terms of flow patterns and used Kattan – Thome –
Favrat maps to verify the flow type in boiling. They concluded that for the boiling
of this refrigerant mixture, nucleate boiling was dominant even at high qualities
with low heat and mass flux values. Also, when there was constant pressure and
flow rate; heat transfer coefficient significantly depended on heat flux.
Ding et al. [30] concentrated on the frictional pressure drop in two-phase flow of a
R410a-oil mixture. They selected 5 mm outside diameter internal spiral grooved
microfin tube to arrange the flow which had 5°C saturation temperature, 200 –
400 kg/m2s mass flux, 7.46 – 14.92 kW/m2 heat flux, 0.1 – 0.8 quality at the inlet,
and 0 – 5 % oil concentration. They revealed that the pressure drop made a peak
at a quality between 0.7 and 0.8 and it was larger for larger mass flux values. In
addition, the presence of oil in the refrigerant increased the pressure drop and
this effect increased with increasing quality values because of the larger
concentration of oil inside the mixture. The authors considered the increase in
pressure drop by comparing with the zero oil condition, and calculated an
enhancement factor to be between 1.0 and 2.2 when compared with the pure
refrigerant. They offered a new correlation with the experimental results using the
local thermodynamics properties of the mixture.
Chen et al. [31] conducted an experiment to examine the two-phase heat transfer
and flow pattern for R134a flow in two ducts. The ducts were narrow, with 1.0 and
2.0 mm gaps and they were placed horizontally. The authors determined the
effects of geometry (duct gap), saturation conditions and quality, heat and mass
flux, on evaporation heat transfer coefficient. These affecting parameters were
between 5 and 15 kW/m2 for heat flux, 0.05 and 0.95 for quality, 5 and 15°C for
saturation temperature, and 300 and 500 kg/m2s for mass flux (for 2.0 mm duct),
500 and 700 kg/m2s for mass flux (for 1.0 mm duct). The results displayed that
heat transfer coefficient was directly proportional (almost linear) with the quality
and at higher mass fluxes the proportionality got steeper. In addition, heat
transfer coefficient increased with increasing heat flux and saturation temperature
and it was larger for the 1.0 mm duct. Yet, these were less effective on this
narrow duct at small heat flux and large mass flux values. The authors also
15
visualized the flow patterns by taking pictures and proposed a correlation for heat
transfer coefficient in two-phase flow of narrow ducts.
Saisorn et al. [32] also studied two-phase R134a flow experimentally in a
minichannel. Their channel was 1.75 mm in diameter (circular), 600 mm long,
horizontal, and made of stainless steel. The results for both flow pattern and heat
transfer coefficient were obtained for the mass flux values from 200 to 1000
kg/m2, heat flux values from 1 to 83 kW/m2 and at 8, 10, and 13 bar saturation
pressures. They calculated the heat transfer coefficient in terms of these varying
parameters and commented that the higher the heat flux, the more was the heat
transfer coefficient, while mass flux and quality did not affect the results
significantly. On the other hand, increasing saturation pressure decreased the
heat transfer coefficient. Moreover, the authors examined five flow patterns for
flow-boiling.
Consolini and Thome [33] attracted attention to the fact that although there is a
common belief and consensus on the higher heat dissipation of micro-channel
flow boiling, the characteristics of this small scale heat transfer still need to be
investigated further. Thus, they studied flow boiling heat transfer of the
refrigerants R134a, R236fa and R245fa in two different microchannels which had
diameters of 510 and 790 µm with mass fluxes changing in the range of 300 –
2000 kg/m2s and with a maximum heat flux of 200 kW/m2. For the two
refrigerants, R134a and R236fa; heat transfer coefficient depended on the heat
flux as well as the fluid properties at not only low qualities but also high qualities,
where as a flow mode, annular flow dominates the flow. Although the third
refrigerant, R245fa, showed a similar behavior as the other refrigerants at low
quality values; at high values, it was independent from heat flux and α-x plane
curves kept increasing for higher qualities.
In light of the literature review presented above, the following general conclusions
can be made:
The saturation pressure has no significant effect on the two-phase heat transfer.
Although for some studies, the increasing saturation pressure increased the heat
16
transfer coefficient drastically, the general consensus was on the negligible effect
on the boiling heat transfer.
The quality change in the two-phase flow was agreed upon for almost all studies
in terms of increasing heat transfer up to moderate quality values, 30 – 50 %. The
change of the heat transfer coefficient for high quality values was concluded in
two ways: continuous increase and decrease after a peak. The studies were
mostly experimental studies and the changes would have occurred due to the
experiment conditions.
Some of the authors in the literature proposed correlations for the numerical and
experiments studies they performed. These correlations are valuable in terms of
comparison and establishing the infrastructure. On the other hand, it should be
noted that the correlations based on the results of these studies would be valid
only for the ranges of the experiments.
The pressure drop for two-phase flow was mentioned to increase in comparison
with the pure liquid flow and it can be commented that a trade-off between the
heat transfer and pressure drop can be expected.
The heat transfer coefficient of two-phase flow of any fluid is considerably larger
in comparison with the single-phase flow. The problems of dry-out, flow
arrangement – vertical and horizontal flow –, the need for high heat flux, which is
the general tendency for the studies to achieve high qualities, could be made as
an inference from the studies.
17
1.3 MOTIVATION AND OBJECTIVES
As it can be observed from the studies in literature, some of which were
presented in the previous section, small scale heat transfer and especially flow
boiling heat transfer in mini- and microchannels is a field that attracts the
attention of many researchers. The results and comments show that despite the
fact that there are several invaluable researches and papers in this area, there is
still need for further investigations. The experimental, numerical, and theoretical
infrastructure about the mini- and microscale heat transfer is being established in
the present era. Among these three main approaches, experimental studies are a
must to increase the understanding, and they lead the numerical and theoretical
studies.
There are basically two approaches used as models in boundary conditions
which are constant wall temperature and constant heat flux approaches. For the
two-phase flow boiling heat transfer in mini- and microchannels, the literature
shows that the studies were focused on mainly the constant heat flux approach. It
is needed to support and extend the constant wall temperature approach, which
is observed in common household refrigerators. On the other hand, when the
mini- and microchannels are compared, because of the difficult manufacturing
and establishment procedures, high pressure drop and no limitation in terms of
space, the microchannel applications would not be practical in refrigerators.
Therefore, minichannels in refrigerators would be a better solution in terms of
increasing heat transfer rate and keeping the pressure drop in a reasonable
amount.
The refrigerant flow in small scale has been widely investigated by several
scientists. Because of the low boiling points of the refrigerants, two-phase flow
studies are also performed to reveal the phase change effect. These studies are
mainly at high pressures because of the fact that the saturation temperatures of
the refrigerants come closer to the room conditions at relatively high pressures
(over 6 – 7 bars). The refrigerants in refrigerators have lower values of saturation
temperatures and pressures, so there is lack of research on the two-phase
18
refrigerant flow in mini- and microchannels at relatively low pressures and
temperatures.
When these three aspects; the value of experimental study, the need for constant
wall temperature approach, and the need for low pressure studies are
considered, in addition to the results of the studies in the literature about two-
phase heat transfer in small-scale media, which show that there is a remarkable
enhancement of heat transfer in small scales; the motivation of this thesis
emerges.
The aim of this thesis is to establish an experimental set-up for investigating the
small scale heat transfer for two-phase refrigerant flow in 1.65 mm copper tube.
The process and progress of the experimental set-up installation is presented
and the design, installation and calibration of the devices, preliminary
experiments, and their results are given.
The following chapter is about the construction of the “Experimental Set-up”, the
preliminary tests and the calibrations of the measurement devices. Chapter 3
focuses on “Two-Phase Flow Experiments” in terms of data analyses and
calculations. The results of the experiments and the comparison with the
available studies and correlations in the literature are given in Chapter 4,
“Experimental Results and Discussion”. The final chapter, Chapter 5
“Conclusions”, concludes the thesis by summarizing the work, commenting on
the conclusions, and making suggestions on the future work.
19
CHAPTER 2
EXPERIMENTAL SET-UP
The aim of this study is to establish an experimental set-up to investigate the two-
phase flow of R134a in 1.65 mm copper tube. The process of the set-up
installation – design of the set-up, equipment used, preparation for the set-up,
calibrations made –, preliminary test and modifications done are given in this
section.
The flow in the experiments is internal laminar flow at steady-state conditions.
The tube used is circular, so the flow can be considered as laminar flow through
a pipe. The experimental results and calculations of the correlations in the
literature are done in the forthcoming Chapter, yet the main theory in terms of
Thermodynamics, Heat Transfer and both General and Particular Laws is
presented in thie following section.
2.1 MAIN THEORY OF THE INVESTIGATION
Heat transfer is defined as the energy transfer study which takes place in the
presence of a temperature difference. There are General Laws – physical laws
independent of the medium – and Particular Laws – that depend on the medium
– that govern the heat transfer study. These laws are very significant for
engineers in all stages of a heat exchange device from design to operation [34].
General Laws for heat transfer are [34]
- the law of conservation of mass
- the first law of thermodynamics
20
- the second law of thermodynamics
- Newton’s second law of motion
Some particular laws about heat transfer are [34]
- Fourier’s law of heat conduction
- Newton’s law of cooling
- Stefan – Boltzmann’s law of radiation
- Newton’s law of viscosity, and so on.
The general forms of the General and Particular Laws for a Control Volume are
[34]:
The law of conservation of mass:
= −.. ∙ .. (1)
The first law of Thermodynamics:
+
+.. + ∙ =..
.. − .. + ∙ .. (2)
The second law of Thermodynamics:
+.. ∙ ≥..
!"#$!.. (3)
Newton’s second law of motion:
% = +.. ∙ .. (4)
21
Fourier’s law of heat conduction:
& = −' ∇ ) (5)
Newton’s law of cooling:
&* = ℎ,)- − )./ (6)
Stefan – Boltzmann’s law of radiation:
&001 = 2)3 (7)
In the interest of this experimental study, the flow is steady, laminar, tubular flow
with one-inlet and one-exit in the continuum regime. The laminar flow is caused
by the low flow rate, which is to be explained in the forthcoming sections. The
verification of the continuum concept is given in the calculations. Radiation heat
transfer is neglected since conduction and convection dominates the heat
transfer process. Therefore the equations mentioned above are simplified due to
the conditions and such assumptions. The simplified versions of the equations
are:
The law of conservation of mass (Continuity equation):
4 = 4 5 = 4 6 (8)
The first law of Thermodynamics:
4 ,ℎ6 − ℎ5/ = .. (9)
The second law of Thermodynamics:
786* = 4 ,6 − 5/ − # 9.:.! (10)
22
Logarithmic Mean Temperature Difference (LMTD) Method:
.. = ; % ∆)=> (11)
Overall Heat Transfer Coefficient:
; = ? @A + BC ,1D 1E⁄ /
GH= IJ (12)
Pressure Drop (applicable only for single-phase flow):
∆K = 4 =$E
MNOP (13)
The main goal in this study is to estimate the heat transfer coefficient and
measure the pressure drop for the two-phase flow experimentally and compare
them with the available studies in the literature. Single-phase laminar flow inside
a circular tube is very-well known since it can be solved analytically; on the other
hand, it is not same for the two-phase flow. The studies in the literature show that
the efforts on the two-phase flow are focused on numerical and experimental
ways to establish a database and construct an infrastructure in this field.
2.2 DESIGN OF THE SET-UP
The set-up is designed to perform two-phase R134a flow experiments. The
necessity of constant wall temperature approach led the design to have two
cycles. Thus, two cycles and a heat exchanger at the evaporation part of the
R134a flow design was done to achieve this goal. These are provided in two sub-
sections “Constant Wall Temperature” and “Heat Exchanger”. In this part the
general design of the set-up is given.
In the early design of the experimental set-up, the test section was decided to be
a glass tube with 1.8 mm internal, 3.0 mm external diameter. With the glass tube,
there could be flow visualization in addition to the experiments. The process
about the change is given in the “Preparation” part.
23
For the two cycles, due to the fact that fluids are needed to be conditioned;
vessels to be considered as reservoirs had to be present to set the fluid
temperature at the designated values. Therefore, two refrigerated circulating
baths – one for each cycle – were put on the set-up to maintain the desired
temperatures.
To circulate the fluids in the system, pumps were used. The aims of the pumps
are to sustain the constant pressure and provide a constant flow rate – with
lowest oscillations possible – at steady-state conditions.
Two-phase flow experiments are the main goal, so the quality of the refrigerant at
the test section must be known. Therefore, a pre-heater was placed just before
the test section to specify and acquire the quality value.
Data acquisition is a must and it can also be considered the most important part
of experimental studies. The flow rates of the fluids that are cooled / heated in the
circulating bath and pumped to system by pumps are to be measured. The
importance of the measurement is significant not only to measure the flow rate
but also to arrange it to the specified values in the experiments. A micro-
flowmeter is selected since the flow rates are relatively much smaller for the
primary cycle if refrigerant cycle is considered as the primary cycle of the system.
For the secondary cycle of the system, two rotameters are placed in the design to
measure the flow rates at the test section and the by-pass. In addition to the flow
rates, the properties of the fluids along the flow at the system pressure and
temperature readings must be recorded. These values are needed for both the
arrangement of the devices to get the desired test points and designation of the
states at the specified locations.
These mentioned parts in the design are located in a schematic of the set-up,
which is presented in Figure 2. The outer cycle in the figure is the refrigerant
cycle and the inner cycle is the water cycle. The fluid used in the secondary cycle
had been thought to be water in the preliminary design but then it was changed
to a mixture to decrease the freezing point. This mixture is composed of water
and ethylene glycol.
24
Figure 2: Schematic view of the set-up at the first design
In the refrigerant cycle starting from the test section exit, refrigerant flows through
the circulating bath to specify the temperature. After the circulating bath, the
temperature at the exit is measured and then there is an accumulation vessel.
This vessel is added to the system to assure the liquid phase of refrigerant before
the pump because pumps operate much more efficiently in liquid phase than in
two-phase or vapor phase. After the vessel, there comes the pump to circulate
the refrigerant inside the system and maintain the steady state conditions in
terms of pressure. The flow rate is measured after the pump by a micro-
flowmeter. The flowmeter should be capable of reading low flow rates, which is
one of the main objectives in the experiments. When the flow leaves the
Pre-Heater
Water Pump
R134a Pump
Micro-Flowmeter
Rotameter R134a Refrigerated
Circulating Bath
Accumulation Vessel
Rotameter
Water Refrigerated Circulating Bath
Minitube Test Section
T T
P T
P
T T
T
∆P
25
flowmeter, the temperature and the pressure of the refrigerant are to be
measured. The reason behind this concept is to determine the state after all the
devices before the pre-heater. Since the state is must be known at least by two
independent properties, the flow must be in single-phase at this location to find
the state correctly. When the state information is acquired, the refrigerant should
be heated up to two-phase for evaporation experiments. There is a pre-heater
placed for this purpose. The name is selected as pre-heater because the
refrigerant is heated just before it enters the test section. The given heat input
from the pre-heater to the refrigerant is also measured to determine the state at
the inlet of the test section. The test section will be described in the “Counter
Flow Heat Exchanger” section.
The water-ethylene glycol cycle is the second cycle of the system. For simplicity,
it is called water cycle / side throughout the thesis. After the test section for the
water side, the fluid comes to the pump. It is pumped and the pressure is
measured at the exit of the pump section. Then the flow is divided into two parts.
One of the divided parts goes to test section whereas the other line goes back to
pump. The reason is to decrease the back pressure effect that can occur in the
tubes or in the test section due to the small diameter. This effect may cause the
pump to operate at very high temperatures and burn-out may occur. After the
division of the flow, the by-pass side – the part that goes back to pump inlet –
starts with a rotameter, which is placed to measure the volumetric flow rate sent
back to the pump. Then the flow is directed to the refrigerated circulating bath of
the secondary cycle. The temperature is measured both at the inlet and exit of
the bath. The temperature is arranged in the bath and then the flow goes to the
junction at the inlet of the pump where it mixes with the flow coming from the test
section. The other part after the division of the flow flows into another rotameter
which measures the volumetric flow rate that goes to the test section. The water
flow at the test section is presented in the “Counter Flow Heat Exchanger”
section.
26
2.2.1 Constant Wall Temperature
Constant wall temperature approach is provided by adding a secondary cycle to
the set-up and flowing another fluid inside this cycle. The aim is to achieve
constant wall temperature at the test section. It is decided to circulate water,
which has a high heat capacity – capable of transferring large amounts of heat in
small temperature differences – and make the flow as fast as possible to increase
the mass flow to decrease the temperature difference. The verification of the
constant wall temperature is needed during the experiments and for this purpose
it is decided to measure the temperature of the wall at the test section from
several locations.
2.2.1 Counter Flow Heat Exchanger
The test section is designed preliminarily as a heat exchanger with counter flow
arrangement. The flow arrangement was selected due to the fact that the
correction factor for the counter flow is 1 (highest value) for phase change on
one-side and it also has higher effectiveness than parallel flow. For the two-
phase flow, counter or parallel flow has no difference, but it is decided to prepare
in this configuration for forthcoming experiments with both single- and two-phase
flow. The heat exchanger is considered as a shell-and-tube heat exchanger with
refrigerant flow on the tube side and water flow on the shell side. Since the flow
of refrigerant is on the tube side, it is advantageous in terms of the heat loss
considerations to the surroundings. The net heat transfer rate from water is the
actual transfer rate to the refrigerant. To calculate the heat transfer coefficient
and pressure drop, the temperatures at the inlet and exit of the test section for
both fluids and the pressure difference between the inlet and exit of the
refrigerant side are measured.
2.3 EQUIPMENT
After the design of the experimental set-up, the devices that are to be used are
specified and obtained. These devices are selected in terms of the needs for the
experimental conditions in addition to the economical sufficiency.
27
As given in the schematic view and mentioned in the design part, a circulating
bath [35], pump [36], and micro-flowmeter [37] are selected for the refrigerant
cycle. The accumulation vessel is provided by Arçelik, the company participating
in the project as the technical supporting partner. For the pre-heater part, heat-
rolled strips were the first design selection. Then the design has been changed
due to the possible problems that may occur in these strips since the geometry is
cylindrical. Therefore, an electric resistance heater was decided to be the pre-
heater of the set-up and a control box that can both manually and automatically
control the heater power input was added to the set-up. This control box was
designed and produced by Yenel Ltd. in Teknokent, METU. The test tube was 1.8
mm inner diameter glass tube in the first design [38] and then it has been
changed to 1.65 mm inner diameter copper tube, which is already present in the
laboratory of the department. For the pressure measurement at the flowmeter exit
of the refrigerant cycle, a pressure transducer is used [39]. For the temperature
measurements, both RTD (Resistance Temperature Detector) probes and
thermocouples are used. An RTD is much more sensitive compared to
thermocouples, so it is used for critical temperature measurements, which are
mainly to determine the state of the refrigerant. On the other hand, there were
two types of RTDs used in the set-up: one has a probe with 0.093” (2.36 mm)
diameter [40] and the other with 1.8 mm. The RTD with the smaller probe is more
sensitive so it was used in the flowmeter exit location. The larger RTD is used to
measure the temperature of the refrigerant at the exit of the bath. Also several
thermocouple measurements are desired along this cycle to verify the RTD
readings. T type thermocouples [41] are selected.
For the water side, a pump, two rotameters, and a circulating bath functioning like
a reservoir are needed. The pump was a 373 W water pump. The refrigerated
circulating bath is the same as that at the refrigerant side [35]. Since the flow at
the water side is designed to be single-phase (liquid) flow, rotameters instead of
flowmeters are used [42]. The temperature and pressure measurements are also
similar to the refrigerant cycle. The temperatures at the inlet and exit of the bath
are measured. The sensitiveness for these points does not have a high
importance so the thicker RTDs are used, which are the same as the primary
28
cycle [40]. The pressure is read by an identical pressure transducer [39] at the
exit of the pump. Thermocouple readings at the water side are also planned, with
the same type.
At the test section, the temperature readings are important for both cycles at both
inlet and exit. The measurements at these points are done by RTDs to have more
accurate results. The critical side is the refrigerant side so the thinner probe
RTDs are used on this side while the thicker probe RTDs are for the water inlet
and exit readings. The pressure drop along the refrigerant flow is another
consideration for the experimental design; thus, a differential pressure transducer
is purchased [43]. The pressure drop for the water side is not of interest.
Moreover, the variation of the wall temperature at the test section is needed to be
measured and thermocouples are to be used for this purpose at several locations
along the refrigerant tube.
For the pressure transducers, DC voltage input is needed so there is a need for a
DC Power Supply [44]. The output of temperature and pressure readings and the
input of pre-heater power values are in terms of voltage and current in volts
(millivolts) and amperes (milliamperes). Since these must be converted into the
necessary units with proper calibrations, it was decided to gather the
experimental data by means of a data acquisition system [45] and transfer them
into a computer interface.
The specifications of these devices mentioned above are given in Appendix A.
These devices are to be connected and mounted on the set-up by tubes, pipes,
and connectors. In the preliminary design, the set-up is drawn in Key-Creator
technical drawing program and length, material, and diameter of the pipes and
tubes; diameter, type, and material of the connectors; and junction and
connection points are determined. This design, which is given as a schematic in
Figure 3, as a sketch in Appendix B, and as tables for connection parts in
Appendix C, was altered during and after the preparation of the set-up.
29
Figure 3: Schematic view of the set-up at preliminary design with states
2.4 PREPARATION OF THE SET-UP
The set-up was prepared with the devices acquired. In this section, the
preparation and establishment of the set-up according to the initial design with
some changes in the water cycle are explained and the schematic view of the
experimental set-up is given in Figure 4.
Minitube test section
30
Figure 4: Schematic view of the set-up at the preparation stage
2.4.1 Structural Frame
Before starting the establishment of experimental set-up, a structural frame was
designed to hold the devices and the system together and in an order. The
technical drawing of this frame is given in Figure 5.
For the frame, 40 mm L brackets were used. In accordance with the technical
drawing, this frame was prepared in METU Mechanical Engineering
Department’s Machine Shop with the permission and support of the professors in
charge. When dimensions of the frame were decided, the subjects like ease of
working, mobility, strength, and space necessity were considered. Since there
Pre-Heater
Water
Pump
R134a
pump
Micro-
Flowmeter
Rotameter R134a Cooling
Bath
Accumulation
Vessel
Rotameter
Water Cooling
Bath
Minitube Test Section
ΔP T T
T
P
T
T
T
T
P
V
Valve
31
are two cycles in the set-up – R134a cycle and water (water + ethylene glycol
mixture) cycle – and since they coincide with each other only at the test section,
two horizontal parts on the frame were designed. Devices on the R134a cycle
and the test section were placed on the upper part and devices on water cycle
were placed on the lower part. Also, the test section was fixed to the vertical part
placed on the upper part. The points that were taken into consideration in the
preparation of the dimensions of the frame are: (a) Length: there should be
enough space to fit into for the devices that forms both cycles separately. (b)
Width: It should be suitable to be transported if there is a need to change the
laboratory. (c) Height: There should be enough distance for the devices to
function properly and the test section should be mounted and/or changed easily
and properly. Moreover, 4 wheels were welded to the bottom of the frame, which
gives the ease and opportunity to move, transport and/or arrange the devices
and set-up itself in short or long term.
After the manufacturing of the structural frame that forms the general structure of
the experimental set-up, 3 pieces of hardboard were placed on the frame to put
the devices on. These hardboards, one placed vertically and two horizontally –
for upper and lower parts,,were 18 mm thick, 2210 mm long, and 800 mm, 640
mm, and 590 mm wide, respectively. They were selected in order to maintain the
entirety of the setup in addition to meet the requirements of the strength to carry
the loads of the devices and other components, the resistance to low
temperatures, the ease of machining and assembly, and to be esthetic as much
as it can be.
32
Figure 5: Technical drawing of the structural frame
2.4.2 Data Acquisition
Figure 6: The photos of the data logger
Before moving on to the fluid cycles and set-up placement, the data acquisition
unit and cards, and the data acquisition process will be described. Initially 2 data
33
acquisition cards were purchased. On each of these cards 20 voltage and 2
current readings can be performed by the slots. In total, out of 40 voltage and 4
current slots, 43 of them were filled in the first preparation of the set-up. All of the
voltage slots were used for temperature readings. Among the 40 temperature
readings, 32 were used for thermocouples, and 7 were used for RTDs. The last
slot was used as a reference for RTD readings and its end was embedded in a
brass block. During this initial establishment process, 3 of 4 current reading
capacities were used and one of them remained empty for the pre-heater control
box and/or other pressure readings that could be added to the system. 2 of the 3
slots used were for compact pressure transducers and the other one was for the
differential pressure transducer. The photos of the data logger are given in Figure
6.
For the temperature readings at the test section, as mentioned in the previous
section, temperature measurements with thermocouples at the surface of the
tube in which refrigerant flows were planned. After the installation of the
experimental set-up, various tubes/channels are desired to be tested. When the
thermocouples are fixed at the surface of the tubes/channels, they will obviously
be shortened by cutting at the time of the change of the test tube/channel. To
prevent this shortening effect to cause any change in data acquisition unit,
connector plugs were decided to be used. These were for T-type thermocouples
[46] and 20 of them with 2 “10 Circuit T-type Standard Panel”s were acquired,
which can be seen in Figure 7 .
In the forthcoming parts, calibrations, placement in the setup, installations, data
acquisition processes, and the aims of the use of the RTDs, the thermocouples,
and the pressure transducers are presented in detail when the establishment of
the experimental set-up is mentioned.
34
Figure 7: The photo of thermocouple plug panels
2.4.3 Refrigerant (R-134a) Cycle
For the refrigerant cycle, the upper horizontal part of the frame structure was
selected. In the refrigerant cycle, 1/4” (1/4 inches, 6.35 mm) outside diameter
copper tubes are used to connect the devices and form the cycle. The material of
the tubes was copper due to the fact that it does not react with the refrigerant
flowing inside and it can resist low temperatures and high pressures. The test
section, which is also part of refrigerant cycle and placed on the upper horizontal
part of the frame, is explained in the “Test Section” part. In this sub-section, the
preparation of the refrigerant cycle from the exit of the test section to the inlet of
test section is presented.
As it is seen in Figure 3, at the test exit of the refrigerant cycle, the temperature
should be measured. For this purpose, a thin RTD and a thermocouple were
placed there. The RTD probe was added to the cycle by a pipe tee, which is also
made of copper and welded by oxygen welding to the copper tubes forming the
cycle. RTD probe was put along the flow to read the temperature value from the
35
fluid whereas the thermocouple was fixed with aluminum tape on the outer wall of
the copper tube, thus it read the temperature at the surface not at the fluid flow.
At the refrigerant cycle, refrigerated circulating baths came after the test section
and temperature measurements. Since the volume of the bath (7 liters) was very
large and the refrigerant coming from the test section was predicted to be two-
phase or superheated vapor, keeping the vapor-phase refrigerant inside the bath
was considered to be impossible due to the possible leakage problems from the
hatch of the bath. Instead, it was decided to prepare a helical coil of the copper
tube to act like a heat exchanger inside the bath. To cool the refrigerant flowing
inside this helical coil, it was needed to fill the bath with another fluid – in liquid
phase –, thus bath would cool this fluid, and it would cool the refrigerant. The
helical coil was prepared longer than it had been planned because of the
possible risks of not cooling the refrigerant enough. It was prepared in the
department’s laboratory, so any modifications could be made easily and quickly.
To provide the cooling of the refrigerant, deionized water and ethylene glycol
mixture was put inside the bath. Under normal conditions, deionized water has a
freezing point of 0ºC and ethylene glycol has a freezing point of -11.3ºC.
However the mixture of these two fluids has a lower freezing point. The bath has
a minimum critical temperature of -44ºC, a minimum operation temperature of -
40ºC, and in the experimental conditions the refrigerant was considered to have
minimum saturation pressure as atmospheric pressure at which the saturation
temperature is -26.3ºC. Therefore, the freezing point of the mixture inside the
bath should be below -40ºC not to freeze, which may cause a decrease in
performance or worse, failure in the bath. Thus, 66% ethylene glycol, 34%
deionized water mixture (volume basis), which has freezing point between -45ºC
and -50ºC [47], was prepared as the bath fluid. The bath and the upper part of
the coil heat exchanger are given in Figure 8.
36
Figure 8: The refrigerant bath and the coil heat exchanger
At the exit of the bath, necessary connections were made to measure the
temperature with both an RTD and a thermocouple, similar to the exit of the test
section. The temperature measurement at this point is very important in terms of
knowing the state of the refrigerant and being able to adjust the temperature to
get the desired values at the other points of the cycle. Therefore, it is invaluable
for the experiments to continue successfully.
The refrigerant is highly evaporative and at vapor state at room conditions, in
addition to the fact that it can mix with air easily. A charge valve was needed to
prevent this mixture with air and fill the refrigerant side with that fluid. The valve
connection was made by pipe tee with welding. The coolest part of the cycle is
the flow inside the bath; therefore, the charge valve was placed at the exit of the
bath just after the RTD. During the experiments, it is desired to know the state of
the refrigerant at specified points, and for the initial flow the phase is an important
parameter to be adjusted. To see whether the fluid is in liquid-phase or not, a
sight glass was added to the system. It was also beneficial to achieve the cooling
in the bath in terms of condensation performance after the refrigerant being
37
charged inside the bath. The sight glass had both side threaded joints and it was
fastened with an internally threaded joint to the copper tube on which a
countersink has been made to fit the sight glass and prevent leakage. On the
other hand; an accumulation vessel was put between RTD and charge valve. The
aim of the use of this vessel was to accumulate the refrigerant cooled in the bath
and provide only liquid flow back to the system. In addition to the sight glass, in a
manner of speaking, they acted as double-check.
It is important to cool and condense the refrigerant inside the bath so the smaller
the volume at the first charge, the faster is the liquid refrigerant. To make the
volume smaller, two on/off valves were put at the inlet and exit of the bath after
the RTD, accumulation vessel, charge valve, and sight glass. The reason of the
selection of the placement was to cool the charged refrigerant, to see the
condensate and to decide the start of the circulation. These valves also had
threaded joints and placed to the cycle with internally threaded joints and
countersink made copper tubes like the sight glass. The section of the refrigerant
cycle from the on/off valve at the inlet of the bath to the on/off valve after the sight
glass is given in Figure 9.
Figure 9: The refrigerant cycle (bath section)
Valve (inlet)
Valve (exit)
Sight glass
Valve (charge)
RTD
Bath
Accumulation Vessel
38
After the sight glass and on/off valve, there comes the gear pump and the micro-
flowmeter in the refrigerant cycle. The pump was placed after the bath exit and
the accumulation vessel to make sure the phase of the fluid coming in was liquid.
It was declared to operate in both phases but it gave better performance in single
fluid phase and also there was the risk for the gears to wear out because the
material of the gears was not steel, instead it was thermoplastic material PPS,
Polyphenylene Sulphide, which has a tendency to wear out in two-phase flow.
The gear pump system was embedded into the system serially and the
connections were made by threaded joints. In order not to have problems with
pump in terms of suction and discharge, mounting straight piping at the inlet and
the exit was taken into consideration. The pump operated at different rotation
speeds and it gave the opportunity to change the mass flow rate without
changing the valve openings, and, without increasing the back pressure effect on
the pump, which heats it up and decreases the operating life.
Figure 10: The refrigerant cycle (the pump and the flowmeter before by-pass line
was added)
39
The micro-flowmeter was also connected to the cycle serially like the pump and
after the pump. It was also fitted with threaded joints and pipes with countersink
on to prevent leakage. On the other hand, valves were added to the inlet and exit
of the flowmeter to make the “zeroing” at the first use and whenever needed. The
use of the valves was strongly recommended by the manufacturer company in
the operation manual and how to operate the zeroing procedure is explained in
the “”Preliminary Tests” section. Moreover, straight pipes at the inlet and exit of
the flowmeter were recommended with a length of 5 – 10 times the diameter of
the pipe at the internal flow in order to minimize the variation of the flow speed. At
the inlet and the exit each, 150 mm straight piping was prepared, which was
much more than 10 times of the pipe outer diameter (6.35 x 10 = 63.5 mm).
Besides, there must be single-phase (liquid) high velocity flow at the startup of
the flowmeter, so a by-pass line was added parallel to the flowmeter to direct the
flow there before opening the flowmeter valves in order to ensure the liquid flow.
The by-pass line was prepared by placing pipe tees before the inlet valve of the
flowmeter and after the exit valve of the flowmeter. These pipe tees were made of
copper as well, and pipes were cut there and they were oxy-welded to the pipes.
Copper pipes coming from both pipe tees were connected at the bottom of the
flowmeter at the by-pass line with a valve to open and close that line when
needed. This valve also had threaded joints. The pictures of pump and flowmeter
(before by-pass line was added) are given in Figure 10 and the picture of
flowmeter with the by-pass line is given in Figure 11.
Figure 11: The photo of flowmeter and by-pass line
40
The refrigerant cycle continued with compact pressure transducer and RTD after
the micro-flowmeter. The state at the exit of the flowmeter must be known in
terms of temperature and pressure so there must be a single-phase (liquid)
refrigerant flowing at that point since pressure and temperature are not
independent properties for two-phase. The RTD connection was made the same
as the other RTDs. The pressure transducer was also connected with a pipe tee
which was oxy-welded to the cycle serially. Two ends of the pipe tee were for the
refrigerant flow, the third one was for the transducer connection. The transducer
had threaded joint as a fitting so a connector part, one end internally threaded
and the other to be welded to the pipe tee, was used to place it to its location.
The refrigerant cycle was completed with the pre-heater after the measurements
before the test section. To provide the ease of use and control, the pre-heater
part was made from a resistance electrical heater which was placed in an
electrically insulated cable and used by winding it around the copper tube. There
were 4 alternatives with 4 different power values they could provide. These were
10 W, 25 W, 60 W and 150 W. The initial calculations showed that 60 W for pre-
heater power would be sufficient, but when the possible future work, with larger
flow rates or different fluids, it was decided to implement the 150 W cable to the
system. The precision of the measurements would be worse with a larger scale
but this problem was decided to be solved by calibrating the heat input to the
refrigerant with the power input to the electrical resistance. The cable with
electrical resistance heater inside that worked as the pre-heater is given in Figure
12 before and after winding.
Figure 12: The pre-heater before and after winding
41
The need for the pre-heater was, as mentioned before, to specify the quality
value at the inlet of the test section. Therefore, it was desired to know the power
input to the system and determine the state, so to control the heat input and
know the value added to the refrigerant, a control unit was decided to acquire.
The previously mentioned control box could control the power input both
manually and automatically. To control the pre-heater from the computer, a
multiplexer card was decided to be purchased, but it was delayed because of
economical and technical problems and a power transducer was used instead.
This power transducer read the power input values from the control box to the
heater and it was transferred to the computer by the data acquisition unit. The
output of the power transducer was in volts but the empty spot for current could
be used by implementing a bridge to be able to read voltage. The control box and
the power transducer photos are given in Figure 13.
Figure 13: The control box and the power transducer of the pre-heater
42
2.4.4 Water (Water + Ethylene Glycol Mixture) Cycle
The second cycle in the system was the one using water and ethylene glycol
mixture as the working fluid. For the simplicity, it is mentioned as “water cycle”
and “water side” within this thesis. In the experiments, the main area of interest
was two-phase flow of the refrigerant; therefore it was needed to try to maintain
constant wall temperature by adjusting the water flow rate. The temperature of
the refrigerant would be almost (since constant temperature at phase change is a
model, and it changes indeed even in small values) constant due to the phase
change at the test section. If this condition can also be achieved for the water
side, the heat transfer coefficient variation with respect to quality change can be
studied for constant wall temperature approach. To minimize the temperature
change of the water at the test section, the mass flow rate of water was designed
to be much more than the flow rate of refrigerant. Also the heat capacity of the
working fluid should be a large value and water satisfies this consideration.
Moreover, the mixture has even larger specific heat value than water. The
mixture composed of the same fluids like in the bath of the refrigerant because of
the fact that the freezing point decreases at a considerable value. The lowest
possible working pressure for the refrigerant was the atmospheric pressure, at
which – as mentioned before – the saturation temperature was -26.3ºC.
Considering that, the mixture was prepared as one-to-one, water and ethylene
glycol, in volume basis [47] and the freezing point of the mixture was calculated
to be approximately between -35ºC and -40ºC. Decreasing the freezing point was
important in terms of making the temperature difference of the two cycles as
small as possible at test section to provide the constant wall temperature
boundary condition. The water cycle had been changed after the initial design in
order to adjust the flow rate and temperature easily. This section is explained in
terms of the new condition of the set-up which was given in Figure 5. As in the
refrigerant cycle, the devices, fittings, and measurement tools are explained in
this section from test section exit to test section inlet. Although it was thought to
have minimum temperature difference between the fluids at the test section,
counter flow was designed in order to increase the heat exchanger performance.
These are presented in “Test Section”.
43
In the water cycle, a pneumatic hose was used to connect the devices and
provide the flow because of the fact that it was easy to make connections,
resistant to working pressure and temperature values, and it did not react with the
working fluid. The geometry of the hose was 6 mm outer, 4 mm inner diameter.
The fitting parts were selected to be compatible with the hose and the device
connections. These fittings were elbows, pipe tees (for pneumatic hose), and one
side threaded, one side hose entrance connectors. One of the most important
aims to use pneumatic hose was that it did not need any welding process as in
the refrigerant side and the length could be changed with respect to the positions
of the devices easily and comfortably. The hose can operate at the temperatures
in the range of -50 – +100ºC so any problem with the flow was not expected.
At the exit of the test section for water cycle; a pipe tee, which had threaded
ends, was used to connect the RTD to measure the water temperature. This RTD
was with a thicker probe and the threaded joints were 1/4”. One of the threaded
ends was connected with a fitting – one end threaded one end hose connection.
After the temperature measurement with the RTD, the water cycle continued with
the circulating bath. The inlet and exit connections of the bath were 1/4” threaded
joints. For the inlet of the bath, a pipe tee was used because both the by-pass
line and the test exit after temperature measurement came to that point. The pipe
tee was with one end threaded, and two ends hose exit. The threaded end was
connected to another pipe tee with three ends threaded. This was done to
measure the temperature of the water mixture at the inlet of the bath. One end of
the second pipe tee was connected to the circulating bath and the other was left
empty for an RTD measurement. Since there was no RTD left for that location it
remained in this stage of preparation. Water and ethylene glycol mixture is not
evaporative like R134a at room conditions and it remains in liquid phase.
Therefore, there was no other medium used to condition the mixture and it
directly flowed into and out of the bath with the desired temperature.
At the exit of the bath the flow was not divided into two as in the inlet. The cycle
continued with the pump after the bath. The temperature was measured at the
exit of the bath with an RTD. It was also mounted by a pipe tee as in the exit of
the test section. The importance of the temperature measurement at the exit of
44
the bath was the ability to know the conditions of the experiment and
performance of the bath by providing the ability to compare the temperature with
the bath temperature.
After the temperature measurement at the exit of the bath, the water cycle
continued with the pump. The connections were threaded joints at both the inlet
and exit of the pump. These joints were one side hose exit and one side
threaded. The pressure of the water was measured after the pump by the
compact pressure transducer of the water side. The pressure transducer was
connected by a pipe tee which had two hose exits and one threaded end used for
the transducer side. The flow was divided into two by a pipe tee after the
pressure measurement. One branch was directed to the test section whereas the
other returned to the bath inlet. For both branches, needle valves were placed in
order to adjust the flow rate. These valves were both side hose connection. The
one in the by-pass part was used to decrease the possible back pressure on the
pump and the one at the test section side was used to adjust the flow rate during
the experiments.
Rotameters were put along the flow serially for each branch after the valves. The
rotameters had both sides threaded joints, so the fittings that had one side
threaded and one side hose were used to connect them to the cycle. They were
selected identical for possible parallel use to increase the flow rate at the test
section. They were placed vertically because this orientation was needed for
proper readings. To place them vertically stable in their position, an L-shaped
aluminum sheet was bent in the department machine shop and the rotameters
were stabilized to that sheet with plastic bracelets. The readings were done
manually since the rotameters were mechanical and had no computer interface
or electronic components.
Afterwards, the flow came to the test section through the hose from the rotameter
and before entering the test section, the temperature was measured with a thick
probe RTD like at the exit of the test section with proper pipe tee and threaded
joints.
45
Figure 14: The water cycle
The hose was totally changed after the initial preparation of the system. The
diameter, 4 mm, was not enough to compensate the flow rate that the rotameter
could read, so the capacity was not used properly. Also the pump reached high
temperatures, which was a risk for burn out or failure. Then, the hose was
changed with another one of the same material having a larger diameter. The
new one had 10 mm outer and 6 mm inner diameter. The picture of the water
cycle with the changed hose and connections completed is given in Figure 14.
Water Pump
Rotameters
Test Inlet
Test Exit
Water Bath
46
2.4.5 Test Section
For the test section in which the experiments would be made, many alternatives
have been considered and a shell-and-tube type heat exchanger was decided to
be built. The designs for the test section for the glass tube at the refrigerant side
were made and until the manufacturing of these designs, a different version was
mounted on the system to perform the leakage and preliminary tests. In this step,
5/8” outer diameter copper tube was used as the shell and 1/8” outer diameter
copper tube was used as the tube side. The tube side was welded to the 1/4”
copper tube that came from the pre-heater exit. After the heater and before
welding the tubes and decreasing the diameter, there were 2 pipe tees placed for
temperature and pressure measurements. The first pipe tee was for the RTD
measurement. The 2 sides for inflow and outflow were welded to the cycle pipes
and the last side was welded to a one end threaded short pipe to connect the
RTD. The second pipe tee was connected to the differential pressure transducer.
All three ends of the pipe tee were welded and third end went to the high
pressure side of the pressure transducer. Since the transducer is differential,
which means that it measures the difference of the pressure between two points,
it was connected parallel to the test section along refrigerant flow. Therefore high
pressure side was joined to the inlet side of the refrigerant at the test section. The
low pressure side (exit) of the differential pressure transducer was welded to the
exit of the test section by a pipe tee.
For both sides of the transducer, threaded joints were used with pipes having
countersink at the ends. The exit of the test section was welded to 1/4” copper
tube before it was welded to the pipe coming from pressure transducer by a pipe
tee. The differential pressure transducer had been used because of the fact that
it directly gave the value of the change in pressure at the test section for the
refrigerant flow. After the pressure transducer low pressure side connection, the
temperature at the exit of the test section was also measured by a thin probe
RTD as in the inlet side. The outer pipe and the inner pipe were connected at the
ends of the test section by welding with copper made covers. There were
temperature measurements for the water side before and after the test section as
well. These measurements were done with RTD probes. In the initial set-up,
47
water side connections were made with the initial hoses (4 mm inner diameter)
and then they were modified as mentioned in the water cycle preparation. The
test section at the first preparation is given in Figure 15.
Figure 15: The photo of the test section at the initial preparation
1- Refrigerant test exit RTD 2- Differential pressure transducer 3- Refrigerant test inlet RTD 4- Refrigerant test exit 5- Refrigerant test inlet 6- Water test inlet 7- Water test inlet RTD 8- Water test exit RTD 9- Water test exit 10- Test section (shell and tube heat exhanger)
The design for the desired test section was composed of the test tube in which
refrigerant will flow and the shell in which the water mixture will flow. The length
of the test section was determined at the initial preparation and the design of the
1
2 3
4
5
8
6
7 9
10
48
shell was made according to it. The shell was made as a combination of identical
parts. They were joined together by bolts and nuts to form the shell and the two
flanges for the ends of the shell were designed in order to close the water side.
The reason of two separate parts in the shell was to be able to get the
temperature readings on the test tube wall. Thermocouples were placed on the
wall at 7 different locations along the test tube and they were planned to cross
the shell and leave it from small intervals left at the connection part of the shell
sides. Measuring the wall temperature was very important for the experiments
because themeasurements would be used to determine the availability of the
constant wall temperature approach, therefore two-phase heat transfer coefficient
for the constant wall temperature approach could be calculated and the variation
of the heat transfer with quality could be revealed, which was one of the
important motivations for the study. The technical drawings of the shell and
flange to be used to form the shell are given in Appendices D.1 and D.2. The
material was selected as aluminum since it was lightweight, easy to machine and
available to be manufactured in short term.
After the leakage and the preliminary tests with this orientation, the test section
was be changed with the glass tube and the shell designed for the experiments.
These tests are explained in the next section.
2.4.6 The Leakage and Preliminary Tests for the Initial Design
2.4.6.1 The Leakage Tests
The most important part after the establishment and before the operation of the
set-up was the leakage tests, which is a must for the proper operation of the
closed-flow systems, and they were performed during the experiments after each
modification or change in the system. These tests were done in the same way for
each and every case.
The leakage tests were performed in two steps. These were tests for the water
side and the refrigerant side, respectively. For the water side, the observation of
any leakage was relatively easier because of the liquid phase of the mixture
49
(water + ethylene glycol). The bath was filled with the mixture and after the
connections were checked, the bath and the pump were started to operate and
the flow was observed. The bath had its own circulation pump; it was needed to
operate the bath together, otherwise it could block the flow and cause the pump
and the liquid to warm up. At some of the junctions, lakage was observed. The
reason was the non-straight cutting of the hose. The flow was stopped and the
hose was either changed or cut straight to prevent leakage. As a result, the
leakage tests for water side were performed and no-leakage condition was
provided with proper arrangements easily.
At the refrigerant side for the initial preparation, there was no need for the flow
during the leakage tests; therefore the pump was not operated. On the other
hand, the refrigerant available was in limited amount, so the leakage tests were
performed by charging air into the cycle. The cycle would operate at clean
conditions; therefore before pressurizing the refrigerant side, vacuum was applied
in order to remove the humidity and possible residues from manufacturing or
welding. After the vacuuming process, the cycle was filled with compressed air up
to 10-12 bars. This value was not beyond the maximum operating pressure of the
devices. The compressor and the charge valve were closed and the pressure
was checked with a pressure gauge. Any pressure loss or larger leaks were
determined easily and were closed by either fastening or welding in terms of their
connection types. Then the system was again pressurized to a high pressure
value. Smaller leaks occurred and the rate of air flowing out of system was
considerably very low. At that point, bubbles made of detergent and water were
used to find the locations of the leakages. It was a long and tiring process to
prevent all leakages, but the no-leakage situation at the refrigerant side was then
accomplished. Afterwards, the system was charged with the refrigerant and both
cycles got ready for the preliminary tests.
2.4.6.2 Data Acquisition Process and the Preliminary Tests
When the leakage tests were completed, the data acquisition and preliminary
tests were prepared. For the data acquisition process after the mounting of the
RTDs, the thermocouples, and the transducers (power and pressure) to the
50
cycles; the connections to the data logger were done. The RTDs and the
thermocouples were attached to the voltage slots and the power and pressure
transducers were attached to the current slots. The calibrations of the
measurement devices were done in Arçelik, and are presented in the
“Calibrations” section. Also the computer interface used for gathering the data
according to the data logger connections and connection types were installed by
Arcelik to collect the experimental values. HP-VEE was used to create the
interface and the flowchart to gather the data. The program was set in order to
acquşre data every 15 seconds and that would be beneficial for steady-state
determination and understanding the response time of the set-up to the different
variables. For the analyses in the experiments, the steady-state conditions were
determined in terms of successful consecutive measurements being in the range
of sensitivity of the devices at least for 50 times (about 10 – 15 minutes).
Before the preliminary tests, the charging of the refrigerant and the filling of the
baths with water and ethylene glycol procedures were done. For the water side,
the mixture was used to decrease the freezing point as mentioned in the previous
parts. During the leakages tests, 7200 gr of deionized water was used in the
water side. Ethylene glycol was not added in that step because water was much
more than ethylene glycol in amount and for the possible leakage losses, this
was not desired. The water was discharged until about 3500 g was present in the
bath. The minimum working temperature for the refrigerant was estimated to be -
26.3ºC, the saturation temperature at atmospheric pressure, so, the volumetric
mixture ratio of 1:1 was selected in order to decrease the freezing point to -37ºC
[47]. The density of the ethylene glycol is 1.123 times the density of water at
room conditions [47], so 3950 gr (± 30 gr) of it was added to the deionized water.
For the refrigerant side, the bath was filled with water and ethylene glycol mixture
as well. The cooling of the refrigerant was provided by the heat exchanger placed
inside the bath as mentioned before. The minimum working temperature of the
bath was -40ºC and with a 2:3 of mixture of water and ethylene glycol
(respectively) in volume, the freezing point of the mixture was decreased to -55ºC
[40]. Then the refrigerant was charged from the charge valve. The valves at the
inlet of the bath and after the sight glass were closed to cool the refrigerant in the
51
bath. The bath was operated and the refrigerant was to become a liquid and
vapor mixture after the cooling. The pressure of the cooled refrigerant decreased
and the charging could go on until the entire refrigerant inside condensed. The
pump was compatible with single and two-phase flow but for the flowmeter, it was
suggested to circulate liquid at high flow rate for at least 15 minutes at the first
operation and also for the experiments, single phase flow was strongly
recommended to get correct and accurate readings. To eliminate any risk of two-
phase flow from the flowmeter, the by-pass line was left open and the inlet and
exit valves of the flowmeter were closed. When the valves next to the bath were
opened, the phase change could be seen from the sight glass that the liquid
refrigerant vaporized because of the instantaneous increase in volume. After the
verification of proper functioning of the data acquisition system and no leakage
situation, the insulations of the pipes, connection points and the test section were
made. The charge of the refrigerant increased the refrigerant side pressure to 5 –
6 bars. High pressure for the refrigerant was useful in terms of achieving liquid
phase (high pressure = high boiling temperature). When there was two-phase
refrigerant present inside the cycle as observed through the sight glass, the
micro-flowmeter inlet and exit valves were closed. At this stage, the pump was
started and the flow followed the by-pass line and the test section before coming
back to the bath.
The water side was also started with the operation of the bath and the pump. The
measurements were checked and when there was no leakage, the temperature
of the bath was adjusted to cool the refrigerant side at the test section to have
liquid refrigerant along the cycle. After the circulation of the water at low
temperatures, the temperature of the refrigerant got low and the liquid amount in
the sight glass increased. When the sight glass was totally filled with liquid, the
flowmeter valves were opened and the system started to work in total. To prevent
the instant changes inside the cycle, the by-pass line was not closed at the same
time, but closed slowly in a while. After that procedure, the pump speed was
increased to maximum to provide 15 minutes high flow rate in liquid phase. Then,
the instructions in the manual of the flowmeter were applied and the zeroing was
performed. The preliminary tests were finished with the analyses of the gathered
data as a first study with the readings.
52
Figure 16: The photo of the set-up after the preliminary tests
The condition of the experimental set-up and the placements of the devices at
this stage are given in Figure 16. Both of the cycles were closed and the flows for
the cycles were performed successfully before the actual experiments.
2.5 MODIFICATIONS
When the preliminary tests were accomplished, the next stage was the change of
test section with the glass tube and the shell made for it. During the preliminary
tests, it was observed that the water pump did not operate properly and it heated
up in the experiments. Therefore a second by-pass line for the water side was
added to decrease the back pressure on the pump. There was a valve on that
section as well, to adjust the flow rate and the pressure of the water side. The
accumulation vessel on the refrigerant side was removed to decrease the volume
R-134a Cycle
Water Cycle
R-134a Bath
Water Bath
Gearpump
Micro-flowmeter
Compressor
Micro-flowmeter
Control Unit
Test Section
Pre-heater section
53
and also it was not useful in terms of liquid sustain to the pump because each
time at the initial operation of the system, two-phase flow was observed. The final
version of the schematic view of the set-up is given in Figure 17.
Figure 17: The schematic view of the set-up in the actual experiments
2.5.1 Ring and Bumper Design for the Glass Tube
When the glass tube and the aluminum shell were mounted on the system, the
connections of the glass tube and the copper tube had been considered in terms
of easy and effective fittings. The copper tubes at the inlet and exit before and
after the test section had 1/8” outer diameter, which were welded to 1/4” outer
diameter copper tubes that formed the cycle. The glass tube had 3 mm outer
diameter which was close to 1/8” (= 3.175 mm), therefore threaded adapter parts
Pre-Heater
Water
Pump
R134a
Pump
Micro-
flowmeter
Rotameter
R134a Cooling
Bath
Rotameter
Water
Cooling Bath
Minitube Test Section
ΔP T T
T
P
T
T
T
T
P
V
Valve
54
were decided to connect glass and copper tubes. These adapters were
composed of 7 parts. One main part in the middle in which the ends of both tubes
entered, two conical parts (one for each side) called “ring” prevented leakage,
two bumper parts (one for each side) protected the ring from torsion and two
head parts (one for each side) joined the main part by threads and fix the tubes
inside the main part to maintain the connection. The ring and the bumper were
the parts that touched the tubes. When the fitting was made, the glass tube broke
because of the metal ring and bumper. The head applied force on these parts to
maintain the connection and prevent leakage but metal parts broke the glass.
Figure 18: The adapter parts for glass - copper transition
The glass tube at the test section was changed with a new one and the problem
occurred again. When the head was fastened the glass tube broke again and
when it was not fastened tight there was leakage. Therefore, it was decided to
manufacture the ring and the bumper parts from a softer material. The technical
drawings for these parts were prepared, which can be seen in Appendix D.3. The
material was selected as polytetrafluoroethylene (Teflon) due to its softness and
machinability. The parts were manufactured according to the technical drawings
in the machine shop of the department. The rings and the bumpers made of
metal and Teflon are presented in Figure 18 and the glass copper transition with
the adapter part is given in Figure 19. The two-part shell was not closed at that
stage in order not to apply force on the glass tube before making copper
Adapter Parts
Metal ring and bumper
Teflon ring and bumper
55
connections. The leakage tests were repeated after the connections and before
closing the water side.
Figure 19: The adapter parts mounted on the cycle
Before closing the shell side, thermocouples were fixed on the glass tube by
pasting with a glue that can operate at low temperatures. To provide no-leakage
on the water side, O-rings were used in the flanges and gaskets in the shell. Thin
openings along the shell at 7 points were made on the gaskets for
thermocouples. After two separate parts of the shell were bolted to each other,
silicone was applied by a silicone gun on the joints, openings and around the
flanges to prevent any possible leakage. This silicone was selected in terms of
pressure and low temperature resistance. The test section with the silicon applied
is given in Figure 20. The photo was taken before insulation so the
thermocouples coming out from the shell can be seen. Red parts on the shell are
the silicone applied regions; blue hoses are water inlet and exit, while black
cables are the RTD cables that go to the data logger. The red silicone is seen on
the flanges too.
56
Figure 20: The test section after the shell was closed with bolts and silicone
2.5.2 Leakage Tests for the Glass Tube
The experiments were to be started after the construction of the test section with
the glass tube and the aluminum shell. When the aluminum shell was closed and
fixed to the set-up frame with silicone, the glass tube could not handle the stress
and it broke as before the adapter parts were used. As the shell side was open
and the flanges were not fixed to the glass tube, the longitudinal and lateral
forces on the glass tube were compensated with the bending of the glass tube.
Yet, after the procedure was followed and water side was closed as well, glass
tube crashed down. For the sake of experiments, the glass tube was changed
with a new one and the procedures were done from the beginning. The shell and
the adapters were taken apart, the silicones were removed and the remaining
was cleaned. The shell was placed on two Plexiglas housings as it is seen in
Figure 20. The height could be adjusted but the lateral and longitudinal effects
could not be handled since it was fixed to the ground with threaded parts and
nuts. The connections were made again and the adapter parts were not fastened
57
so tight not to break the glass again. But when it was not tight, there was leakage
because of the gas phase and high pressure of the refrigerant. The adapters
were become tighter and there was a considerable effort in order not to break the
glass tube again. When the no-leakage condition was satisfied, the insulation of
the test section was needed. In any small touch while covering the test section
with insulation material, the glass tube kept on breaking. For three more times the
procedure was repeated but every time the glass was broken at least in the
insulation process. At the last time, the glass was broken at the end of it, where it
entered the adapter; then, it was decided to fill the adapter part with glue which
was resistant to refrigerant, low temperatures and high pressures. A cylindrical
mold to hold the glue was prepared and the adapter and the joints were filled with
the glue. The leakage was decreased but it was observed that the rate of
decrease in pressure increased in time. The pasting operation was also repeated
again and again but the leakage problem could not be solved. The photo of the
solidified glue can be seen in Figure 21.
Figure 21: The exit of the test section with solidified glue at the adapter
58
For the closed system of both sides, the experiments could not be conducted
when there was leakage. Otherwise, the results would have deviated significantly
because of the improper measurement of mass flow rate. The glass tube had
been selected for the possibility of flow visualization; however, since the main aim
of the set-up was to investigate multiport minichannels after the verification of the
experimental set-up, the test section was decided to be change, as explained in
the next section.
2.5.3 Copper Tube as a Replacement for the Glass Tube
The leakage problem with the glass tube could not be solved. Since the problems
occurred due to the brittle structure of the glass material the most suitable
replacement for the glass tube was considered as copper tube having 1/8” outer
diameter. The threaded joints were one of the sources of leakage as well, so 1/4”
diameter copper tube was decreased to 1/8” diameter copper tube before the test
section.
Figure 22: The test section with the copper tube
59
The internal diameter for the glass tube was 1.8 mm and the copper tube had
1.65 mm outer diameter. Therefore the change wwould not affect the heat and
mass flux in a significant amount. The outer diameter change caused the change
in the holes of the flanges. These holes were widened from 3 mm to 3.2 mm.
Thus, the copper tube with 1.65 mm internal diameter and 3.065 mm outer
diameter – about 1/8” (=3.175 mm) – were put to the test section. The test
section after the replacement is presented in Figure 22.
The connections of the cooper tube were accomplished with welding it to the
same diameter copper tube at the inlet and exit of the test section. The water side
was the same – two-part shell and two flanges – and the no leakage condition
was achieved easily with welding. The insulation of the test section and the tubes
were made and the system became ready for the experiments. The insulated test
section is given in Figure 23.
Figure 23: Test section with copper tube after insulation
60
2.6 CALIBRATIONS
The calibration of the components in the set-up can be investigated in two main
parts: Calibrations of the measurement devices and calibrations of the losses
from the parts that affected the experimental results.
The calibration of the micro-flowmeter had been done before the purchase by the
company and the mass flow rate of the refrigerant was read from the transmitter
unit. The transfer of this data to the computer could not be done because it
needed a special card on the data acquisition unit, and the card was not acquired
due to the economical reasons. The value was collected manually from the
transmitter in unit of g/s. The rotameters were mechanical and they were
calibrated default for liquid flow. The volumetric flow rate was read manually from
the rotameters in l/min. The other measurement devices were the pressure
measurement devices and the temperature measurement devices.
The losses from the system to the surroundings were important in terms of
having proper readings and measurements to calculate the desired quantities
correctly. There would be losses from all the components and pipes along the
cycles. On the other hand, only the losses that affect the test section were taken
into consideration since they designate the experimental results. These losses
were the losses from the pre-heater and the losses from the test section. The
pre-heater gave heat power input to the refrigerant but the given power could not
be transferred to the refrigerant with 100% efficiency. The input power was very
important in terms of finding the quality of the refrigerant at the inlet of the test
section, thus the calibration of the pre-heater in terms of losses were done. The
losses from the test section were calculated in terms of two different components.
In total it was the loss from the water to the surroundings, since the refrigerant
flowed inside the tube, the loss was from water side. It was composed of the loss
and the viscous heating. In the subsections, these calibrations are explained in
more detail.
61
2.6.1 RTD and Thermocouple Calibrations
The calibrations of the temperature measurement devices were done in Arçelik
laboratories [48]. This subsection was prepared based on their study. For the
calibration of thermocouples and RTDs, a cooling bath and a reference
temperature measurement device (a reference RTD) were needed. The
thermocouples and RTDs, which were to be calibrated, were put in the bath with
the reference RTD. The temperature measurement data from all the devices
were taken with varying bath temperatures. Based on the results of the reference
RTD, the calibration constants for thermocouples and RTDs were found with
curve fitting methods. The temperature intervals in the calibration should be
determined according to the operation temperature range and the type of the
temperature measurement device. The calibration values should be scanned in
small temperature intervals at the operation temperature range and it can have
larger temperature difference out of the range. The temperature curves for all
RTDs and thermocouples were acquired. The curve fitting for all was linear and
the constants of these calibrations were installed to the computer program. A
sample calibration curve for an RTD is given in Appendix E.1.
2.6.2 Transducer Calibrations
The transducers used in the experiments were one differential pressure
transducer, two compact pressure transducers and one power transducer.
The power transducer gave 0 – 10 V voltage output and measured 0 – 1000 W
power values. This transducer was connected to the control box that adjusts the
power input to the pre-heater. The output of the power transducer was connected
to the empty current slot in the data logger cards and a bridge resistance was
used to convert the voltage output to current so it could be read by the data
logger and transferred to the computer program. The connections and the
modifications in the reader card were done by Arçelik. The voltage to watt
constant for the power transducer was 100 and it was written to the program as
an input to read the correct values. The accuracy of this value was not important,
because fırther calibrations were made for the pre-heater.
62
The calibrations of the pressure transducers were also performed by Arçelik in
their laboratories in Istanbul. The calibrations for compact pressure transducers
(CPT) were done together since they were identical and had the same range. A
reference measurement device, reference pressure transducer (PT) was used
and the CPTs were connected to a system in parallel with the reference PT. The
pressure in the system (pressured nitrogen gas in a closed pipe with valves to
control the pressure of the system) was varied from 0 to 200 psi which was the
range of the CPTs. The variation of the pressure was first increasing from 0 to
200 psi and then decreasing from 200 to 0 psi. The outputs of the CPTs in the
unit of mA were recorded for both of them, separately – although they were
identical CPTs, each had to be calibrated. In the calibration process, 1 psi
increments were used for the range of 0 – 15 psi and 5 psi increments were used
for 15 – 200 psi values. The output of the CPTs were 4 – 20 mA and the pressure
values corresponding to the output milliamperes were drawn and curve fitting was
applied to designate the constants. For the two transducers, which would be used
in refrigerant and water sides to measure the pressure, the following equations,
for CPT 1 and CPT 2 were acquired, respectively.
P [psi] = 12682 I [mA] – 50.417 (14)
P [psi] = 12705 I [mA] – 50.520 (15)
In addition to the CPTs, the calibration of the differential pressure transducer
(DPT) was accomplished, which would be used to determine the pressure drop at
the test section. The range for the DPT was 0 – 5 psi (= 0 – 0.34 bar) so the
calibrations were performed for this range. The reference PT that calibrated the
CPTs had one exit, whereas the DPT had two sides, high pressure and low
pressure sides. So while performing the calibrations, the high pressure end of the
DPT was connected parallel to the reference PT and the other end was released
open to the atmosphere. For the first value of the DPT, which was “0” bar, the
high pressure side was not pressurized and the current output corresponding to
this “0” pressure difference condition was recorded. Then for the increasing
temperature of the high pressure side, the output current values were recorded
63
corresponding to the pressure values read. The effect of the atmospheric
pressure was subtracted from the pressure value read at the reference PT so the
calibration of the DPT was successfully completed. Curve fitting to the calibration
results was also applied as in the CPTs and the following equation was obtained.
DP [bar] = 21.452 I [mA] – 0.0878 (16)
The calibration results and the calibration curves for the pressure transducers are
given in Appendices E.2 – E.6.
2.6.3 Heater Loss Calibrations
The heater (pre-heater), was placed before the test section for two-phase
experiments. Before explaining the calibration of the pre-heater, its purpose
should be mentioned again. The temperature and the pressure of the refrigerant
are measured at the exit of micro-flowmeter and for the state determination,
these properties should be independent; therefore, at that location the refrigerant
should be in liquid phase. For the two-phase flow, pre-heating was decided to be
applied. To a known state, heater power is induced and the enthalpy of the
refrigerant was increased until the desired quality was achieved. The quality
designation was a significant process and there should be as much accuracy as
possible. The pre-heater power, which was electrical resistance and wounded
around the copper pipe before the test section, was selected as 150 W and a
control unit was prepared to adjust the input power manually and automatically.
In the experiments, the automatic control was not activated due to the technical
problems of the related data logger card. The manually adjusted power input was
transferred to the data logger by a power transducer and it could be read from the
computer program. After the pre-heater was mounted on the system, it was
covered with insulation material (Armaflex flexible thermal insulation made from
Elastomeric foam based on synthetic rubber [49]). The thickness of the insulation
material was 6 mm with a thermal conductivity of 0.040 W/mK. For the verification
of possible heat losses, three thermocouples were placed along the pre-heater.
One thermocouple was put on the copper pipe under pre-heater, one between
the pre-heater cable and insulation – on the pre-heater, and one on the
insulation. When the preliminary tests were being conducted, it was observed
64
that there was temperature difference between the two faces of the insulation,
and the outer face of the insulation was at a higher temperature than the
surroundings temperature, which meant there was loss from the heater section.
The calibration of the pre-heater was done with single phase flow of the
refrigerant. The temperature and the pressure of the refrigerant was measured at
the inlet of the pre-heater and to calculate the power transferred to the refrigerant
from the pre-heater, the state of the refrigerant should be determined at the exit
of the pre-heater. The temperature was measured at that position and for the
state determination at the exit of the pre-heater, single-phase flow was needed.
When the control volume analysis was performed at the pre-heater section for the
refrigerant, there was one inlet and one exit, no work, and positive heat transfer
along the control volume as it is seen in Figure 24.
Figure 24: The pre-heater section of refrigerant cycle as control volume
The continuity equation for the refrigerant cycle at the pre-heater section is Eq.
(8), since there was one inlet and one exit, with constant mass flow rate. The first
law of thermodynamics for the control volume is given below.
QR = 4 Q,ℎ@6 − ℎ@5/ (17)
T T , P
Inlet Pre-heater Exit
4
65
In this equation; QR is the heat transfer rate as input to the refrigerant at the pre-
heater section, 4 Q is the mass flow rate of the refrigerant, ℎ@6 is the enthalpy of
the refrigerant at the pre-heater exit, and ℎ@5 is the enthalpy of the refrigerant at
the pre-heater inlet. The enthalpies of the refrigerant at the mentioned locations
were determined with the pressure and temperature values. The phase of the
fluid was adjusted by the pump and the bath to be liquid. The pressure at the exit
of the pre-heater was assumed the same as the inlet because the pressure loss
from the inlet to exit was calculated as negligible.
Table 1: The states for the calibrations of the pre-heater
Name Unit Experiment 1 Experiment 2 Experiment 3 Mass flow rate g/s 1.50 2.30 1.96 R134a pressure bar 4.61 3.39 3.61 Pre-heater inlet ºC 4.70 -2.55 -2.07 Inlet enthalpy kJ/kg 205.92 196.45 197.29 Pre-heater exit ºC 12.58 -1.20 2.57 Exit enthalpy kJ/kg 217.15 198.64 203.90 Pre-heater input W 19.11 9.59 15.71 Pre-heater transferred W 16.83 5.03 12.94
Name Unit Experiment 4 Experiment 5 Experiment 6 Mass flow rate g/s 1.86 1.67 1.94 R134a pressure bar 4.72 4.42 3.79 Pre-heater inlet ºC -0.27 -0.51 -4.97 Inlet enthalpy kJ/kg 199.70 199.34 193.47 Pre-heater exit ºC 13.66 10.91 5.70 Exit enthalpy kJ/kg 218.96 215.14 207.72 Pre-heater input W 39.13 32.75 32.81 Pre-heater transferred W 35.78 26.34 27.65
The mass flow rate was read from the monitor of the micro-flowmeter. When the
oscillations were in a small range, 50 consecutive values were recorded and the
average was taken like in the other measurements gathered by the data logger.
When the enthalpies and the mass flow rate for the pre-heater section were
determined, the power input (heat transfer rate) to the refrigerant could be found.
The pre-heater power input could be adjusted manually from the control box and
the value was transferred to the computer program by the power transducer. For
different power input values; the mass flow rate, and the enthalpies were
66
designated and the graph of the heat transfer rate transferred to the refrigerant at
the pre-heater section versus power input from the pre-heater was drawn to find
a relation to be used in the experiments for two-phase flow. The states for six
pre-heater calibration experiments are given in Table 1.
In the given table, the previously mentioned properties, power input value and
mass flow rate were measured for each experiment and the necessary properties
were found from the thermodynamic properties of R-134a. The “Pre-heater input”
value stands for the power transducer reading value from the computer program,
while the “Pre-heater transferred” value was the calculated heat transfer rate (Eq.
(17)). These values, which were used to calculate the pre-heater calibration
curve, are given in Table 2.
Table 2: The pre-heater calibration curve values
Pre-heater Input Pre-heater Transferred W W
9.59 5.03 15.71 12.94 19.11 16.83 32.75 26.34 32.81 27.65 39.13 35.78
The calibration curve is presented in Fig. 25. This curve was used for the two-
phase experiments for the determination of pre-heater power input values. The
pre-heater calibration experiments were conducted at 3 – 5 bar refrigerant
pressure values and the two-phase experiments were made at 4 – 7 bar pressure
values. The saturation pressure dependence in two-phase experiments were
concluded to be insignificant, so the difference had no important effects. On the
other hand, the two-phase flow experiments were made at saturation
temperatures, whereas the pre-heater calibration experiments were conducted
for single-phase flow. The calibration experiments were at temperatures close to
saturation at the pre-heater exit and the room temperature was between 20 – 22
ºC during the experiments (the difference was due to seasonal change and it did
67
not cause any considerable effect on the results). The correlation in the
calibration curve came out as:
QR = 0.9534 W − 2.9294 (18)
Figure 25: The calibration curve of the pre-heater
2.6.4 Test Section Loss Calibrations
In the test section, there was counter-flow shell-and-tube heat exchanger with
one shell and one tube pass constructed as mentioned before. In the tube side,
y = 0,9534x - 2,9294R² = 0,9828
0
5
10
15
20
25
30
35
40
0 5 10 15 20 25 30 35 40
QR
I(W
)
QH (W)
68
refrigerant flowed and in the shell side there was water and ethylene glycol
mixture. The calculation of the heat transfer rate from the water side to the
refrigerant side could not be done using the refrigerant temperature and pressure
values, since two-phase flow was aimed in the refrigerant side and the
temperature was not an independent property from pressure. Therefore, it was to
be calculated from the water side measurements and calculations. The water
side was covered with insulation to minimize the losses but if the losses had been
assumed to be negligible, the results would have deviated in a significant
magnitude. The losses from the test section – in other words, from the water side
– was to be found. These losses were categorized in two ways: viscous heating
and the loss to the surroundings. The heat transfer rate from the water side to the
refrigerant side, -, was:
- = Y ∙ Y,ℎY5 − ℎY6/ + − -,[\]] (19)
In this formula, Y is the volumetric flow rate of the water, Y is the density of the
water, ℎY5 and ℎY6 are the water side inlet and exit enthalpies, is the viscous
heating, and -,[\]] is the water side heat loss to the surroundings.
For the calculation of the difference of the enthalpies at the inlet and exit of the
water side in the test section, the equation below was used due to the fact that
the water mixture was liquid.
ℎY5 − ℎY6 = ^_ ,)Y5 − )Y6/ (20)
Here )Y5 and )Y6 are the temperatures of the water at the inlet and exit of the test
section, respectively.
The viscous heating is an important consideration in low velocity and low
Reynolds number flows of very viscous fluids [50]. Before calculating the viscous
heating value, the geometry of the shell side is presented in Table 3. The
technical drawing is also provided in Appendix D.2. Inner diameter is the outer
diameter of the copper tube and the outer diameter is the inner diameter of the
69
shell. The cross sectional area is the water flow area and the hydraulic diameter
is the ratio of 4 times the cross sectional area over the wetted perimeter (the sum
of inner and outer perimeter).
Table 3: Water side geometrical constraints
Inner diameter (mm) 3.18 Outer diameter (mm) 10.00
Inner perimeter (mm) 9.97
Outer perimeter (mm) 31.42
Cross sectional area (mm2) 70.62 Hydraulic diameter (mm) 6.83
For the different temperatures of water and ethylene glycol mixture, the
properties were also needed for the calculations of the calibrations and the
experiments. The properties of the mixture for different mixture ratios are
presented for different temperature values [40]. The behaviors are not given and
the temperature intervals are relatively large for accurate property readings.
Therefore, with the use of these data points as reference, property curves for
water and ethylene glycol 1:1 ratio volume basis were prepared. These properties
are specific gravity (the ratio of the density of the mixture over the density of the
pure water), specific heat, dynamic viscosity, and thermal conductivity.
The properties were taken as functions of temperature only, because the mixture
was assumed to be incompressible. The data points of the properties and the
corresponding temperature values are presented in Appendices F.1 – F.4 [47].
Among these four properties given in the tables, specific gravity, specific heat,
and thermal conductivity showed a linear dependence on the temperature, while
dynamic viscosity had a second order function with temperature. The graphs and
the curve fittings with the equations are also presented in Appendices F.1 – F.4.
In the curve fittings, “y” is the property in that graph and “x” is the temperature.
The relations were also found for the temperature in degrees Celcius (°C).
70
These properties for the water mixture were used in all calculations of the water
side. They were not given in the two-phase flow experiments or the results
section since the same tables and graphs were used to determine the properties.
For the designation of the viscous heating on the water side, the temperatures of
the water side and the refrigerant side were equalized with the temperature of the
surroundings. The temperature of the water and refrigerant sides were controlled
from the inlet and exit RTDs, and the surrounding temperature was followed from
the thermocouples that were open to the surroundings. For these conditions, the
water side was operated with different flow rates to see the effect of Reynolds
number (Re) on the viscous heating. Several experiments were conducted and 4
successful readings were taken to make the calculations. Although the number of
successful readings was seemed to be not enough, the same readings were
recorded for the same Re values on water side in 5 – 10 % range. On the other
hand, the dependence of viscous heating on Re was observed to be linear
showing that the results were sufficient. The viscous heating value was found in
terms of watts (W) and the table, the graph, and the correlation are presented in
Table 4, Figure 26, and Eq. (21), respectively.
Table 4: Viscous heating experimental values
Unit Exp. 1 Exp. 2 Exp. 3 Exp. 4
Pwater bar 13.20 12.77 13.06 12.63
Vwater l/min 1.15 1.85 1.50 1.60
Twater, mean °C 20.01 19.86 20.36 19.83
ρwater kg/m3 1080.59 1080.67 1080.42 1080.68
mwater kg/s 0.0207 0.0333 0.0270 0.0288
Cp J/kgK 3387.21 3386.70 3388.35 3386.61 µ kg/ms 0.00268 0.00269 0.00266 0.00269 k W/mK 0.403 0.403 0.403 0.403 V m/s 0.271 0.437 0.354 0.378 Re - 747.81 1198.93 981.97 1036.34
QV W 9.69 14.50 12.55 11.91
Twater - TR-134a °C -0.19 -0.52 -0.20 -0.18
71
Figure 26: Viscous heating versus Reynolds number
[] = 0.0103 ∙ cdYe61 + 1.9879 (21)
Besides the viscous heating, there is heat loss to the surroundings through the
insulation and the edges of the heat exchanger. The dependence of the heat loss
on the temperature difference between the water side and the surroundings was
to be determined. The water side temperature was 35°C maximum and the
ambient temperature in the experiment room was 20°C minimum. The loss was
calculated from the equation (22) for these extreme temperature values.
-,[\]][] = h!i
jklEmnopqrsklD klE⁄ tPjuvqm pqrskErl klD⁄ t
PjuErlm p ijkErlmnwEx
(22)
In this equation, the first term in the denominator is the convection resistance for
the water side, and hW is the heat transfer coefficient of the water. The second
and the third terms are the conduction resistances along the aluminum shell and
8
10
12
14
16
700 900 1100 1300
Vis
cous
Hea
ting
(W)
Re
72
the insulation. The last term is the convection resistance on the air side
(ambient), since the temperature of the room is recorded.
The water side heat transfer coefficient was calculated using four different
equations. The correlations that were used to determine the heat transfer
coefficient of the water are given in equations (23-26) below. The thermally
developing flow correlations were taken for all experiments.
Correlation of Gnielinski [52]:
y = ?3.66 + 1.61 |6$= I ⁄
(23)
valid for 0.1 < |6$= < 10000.
Correlation of Hausen [52]:
y = 3.66 + .,|6$ =⁄ /.p.,|6$ =⁄ /. (24)
also valid for 0.1 < |6$= < 10000.
“Asymptotic mean Nusselt number in circular ducts” [52]:
y = 1.953 |6$= ⁄
(25)
valid for |6$
= > 100.
Correlation of Pohlhausen [52]:
y = 0.664 |6$= ⁄ K0J ⁄ (26)
valid for 0.5 < K0 < 500 & |6$= > 1000.
73
In these equations, Nu is the Nusselt number, Pe is the Peclet number, d is the
hydraulic diameter and L is the length of the shell. Pe is the multiplication of Re
(Reynolds number) and Pr (Prandtl number).
The water side geometry given in Table 5 was used in the viscous heating
calibrations. For the heat loss to the surroundings, some of the other necessary
geometrical parameters are also presented in Table 5.
Table 5: The water side geometry for heat loss calculations
Inner diameter (mm) 3.175 Outer diameter (mm) 10 Inner perimeter (mm) 9.97 Outer perimeter (mm) 31.42
Cross sectional area (mm2) 70.62 Hydraulic diameter (mm) 6.83 Equivalent diameter (mm) 28.32 Test length (mm) 400
The temperature dependence of the loss to the surroundings was to be found,
but after the calculations for the maximum temperature difference and the
maximum heat transfer coefficient for the water side, it was decided to neglect
this term. The geometrical parameters, thermal conductivity, and results for the
heat loss are given in Table 6.
The heat loss came out to be 1.22 W for the maximum case. The heat transfer
coefficient for the water side was selected as the maximum of the results of
equations (23-26). The heat transfer to the refrigerant for the maximum case was
about 50 W, so the maximum loss was about 2.5%. Also the situation was
investigated for the minimum heat transfer and maximum losses, and the result
was 0.25 W heat loss for 9.5 W heat transfer rate. So the loss was again about
2.5%. In order to ease the calculations, the term was neglected in the
experimental calculations.
74
Table 6: The results for the maximum heat loss to the surroundings
dsi mm 10
dso mm 20
tins mm 20
dins mm 60
kal W/mK 230
kins W/mK 0.04
hW W/m²K 1000
hair W/m²K 10
L mm 400
∆T K 15
Q W 1.22
75
CHAPTER 3
TWO-PHASE FLOW EXPERIMENTS
After the construction of the set-up, connecting the devices, making preliminary
experiments and calculating the losses to the surroundings; the next step was the
two-phase flow experiments. This Chapter is divided into 5 sections: Approach,
Conditions, Flow Chart, Data Analysis, and Calculations.
3.1 APPROACH
Two-phase flow of refrigerant inside the test section was desired. The studies in
the literature were focused on the constant heat flux approach. Moreover, the
mass flux values were relatively large for the refrigeration processes. The main
goals in the two-phase flow were to achieve these conditions; constant wall
temperature and low mass fow rates. The quality and the heat flux were
mentioned to be the most dominant parameters on the two-phase heat transfer
coefficient. Therefore, the quality variation was an important consideration in the
investigation. On the other hand, mass flux, saturation temperature and pressure,
and the pressure drop were the other parameters to be examined.
The quality of the refrigerant could not be measured like pressure and
temperature. Moreover, the temperature and pressure are not independent
properties for phase-change. Therefore another independent property is needed
to determine the state for a two-phase location. The pressure was selected as
one of the properties to designate the state, since the measured temperature
fluctuates for a two-phase flow; the constant temperature approach during phase-
change is an approximation [51]. In fact, the temperature tends to show an
76
oscillatory behavior and the average value is taken to find the saturation
temperature. The refrigerant was pre-heated and the single-phase refrigerant
became two-phase before the test section. The quality right after the pre-heater
was taken as the inlet quality, and along the test section, the constant wall
temperature approach was applied and the heat transfer was given from water
side to the refrigerant side. This heat transfer rate was calculated from the heat
loss and viscous heating correlations (Eq. (19)) and the energy of the refrigerant
was increased.
The quality and enthalpy were known at the inlet. Adding the heat transfer rate to
the inlet conditions and applying the first law gave the exit enthalpy. The pressure
drop was measured at the test section, so the exit pressure was known for the
refrigerant side. Combining the exit enthalpy and pressure, the state at the exit
was determined. The adjustments of the test section, flow rate, temperature and
pressure values to get successful readings and the data analysis are presented
in the “Flow Chart” and “Data Analysis” sections.
3.2 CONDITIONS
The two-phase flow experiments were conducted for varying conditions. These
parameters can be divided into three parts: Refrigerant side, water side and the
pre-heater power input.
The refrigerant side conditions are given in Table 7. The properties of the
refrigerant were read from the program [54] and the graphs were prepared
according to the saturation pressure [54], given in Appendix G. The conditions for
the two-phase flow that changed in the test section were the mass flow rate, the
inlet and exit pressure and temperature, the inlet and exit quality, and the
pressure drop.
Table 7 shows that the mass flow rate of the refrigerant was between 0.34 and
0.94 g/s. These values were selected according to the operation flow rates of the
refrigerants in actual refrigerators. The range for the refrigerator operations is 0.2
– 1.0 g/s. The maximum value could be satisfied but the flowmeter problems did
77
not let successful readings for the minimum flow rate experiments. Yet, the
experiments could be considered sufficient in terms of the mass flow rate range
since the majority of the range was investigated.
Table 7: The refrigerant conditions for the test section in two-phase flow
Parameter unit minimum maximum Refrigerant mass flow rate g/s 0.34 0.94 Refrigerant inlet pressure bar 4.59 7.76 Refrigerant inlet temperature ºC 13.08 30.26 Refrigerant inlet quality - 1.20% 70.16% Refrigerant pressure drop bar 0.02 0.21 Refrigerant exit pressure bar 4.38 7.71 Refrigerant exit temperature ºC 11.65 30.03 Refrigerant exit quality - 16.57% 95.68%
The inlet pressure values to the test section were relatively large for the
refrigerator applications. The reason was that the condensation of the refrigerant
at low pressures was a problem in terms of the cooling bath performance for the
refrigerant side. The bath temperature could decrease to -35 – -40ºC but the
refrigerant temperature could not be decreased to these values since there was a
coil heat exchanger used instead of circulating the refrigerant inside the bath. On
the other hand, the studies in the literature showed that for the two phase
experiments, the saturation pressure effects were almost negligible. Therefore,
the experiments were made with inlet pressure values of 4.5 – 8 bars. The
pressure drop was between 0.02 – 0.21 bar (0.29 – 4.58 %, the
minimum/maximum pressure drops were not for the minimum/maximum inlet
pressures), so there was no significant difference between the inlet and exit
pressure values. Therefore, the saturation pressure was taken as the average of
the inlet and exit pressures.
The quality values of the refrigerant at the test section inlet was changed from
1.20% to 70.16% and the exit was from 16.57% - 95.68%. When calculating the
heat transfer coefficient, the average of the inlet and exit was taken as the quality
value for the test section, similar to analyses in literature. The average quality
78
increase in the test section was 20%, so taking the average quality value as the
experimental parameter while presenting the results was acceptable.
Considering the average pressure of the refrigerant as the saturation pressure
and the average quality as the quality of the experiment, Table 8 shows the two-
phase experimental conditions for the refrigerant in terms of mentioned
parameters.
Table 8: The average refrigerant side parameters for the two-phase experiments
Parameter unit minimum maximum Refrigerant saturation pressure bar 4.49 7.73 Refrigerant saturation temperature ºC 12.41 30.13 Refrigerant average quality - 8.29% 82.92%
Based on the change in the refrigerant properties, water side was adjusted in
terms of flow rate and bath temperature. The pressure on the water side was
about 10-12 bars because of the pump power and the small diameter pneumatic
hose. However, the pressure change did not affect the properties of the water
since the mixture was in liquid phase and almost incompressible. Water side
parameter ranges during the two-phase flow experiments are given in Table 9.
Table 9: The water side parameter ranges for the two-phase flow experiments
Parameter unit minimum maximum Water volumetric flow rate l/min 0.3 2.1 Water mass flow rate g/s 5.37 37.62 Water Reynolds number - 136.94 1333.54 Water inlet temperature ºC 14.6 34.37 Water exit temperature ºC 14.45 34.32
The flow rate on the water side was adjusted from the valves and the flow rate
was manually read from the rotameter. The range of the rotameter was 0 – 2.1
l/min, and the highest value was used in the high flow of the refrigerant. The
temperature of the water was adapted to the refrigerant temperature to get
constant wall temperature at the test section. When the water side losses were
79
calculated, the correlations for developing laminar flow were selected. From the
table, it can be seen that even for the highest flow rate value, the Re value did
not exceed 1335. Moreover, the lowest value of water mass flow rate was about
6 times the highest flow rate of the refrigerant, and the average mass flow rate of
the water was 33 times the average mass flow rate of refrigerant. The reason
was that the higher the water mass flow rate, the closer was the constant wall
temperature approach.
Besides the parameters of the two cycles, the heat transfer rates were varied due
to the pre-heater power input and Re of the water. The ranges for the pre-heater
section and the losses from the test section, in addition to the heat transfer rate
from the water side to the refrigerant side are given in Table 10.
Table 10: The pre-heater and the test section heat transfer rates for two-phase
experiments
Parameter unit minimum maximum Pre-heater power input W 3.6 57.18 Pre-heater net power transferred W 0.50 51.59 Viscous heating W 3.40 15.72 Heat transfer rate to the refrigerant W 9.70 49.91
The pre-heater power was varied to get different inlet quality values for the
refrigerant. It was seen that the refrigerant was pre-heated as much as it was
evaporated at the test section. The importance of the calibrations can also be
verified from the comparison of the pre-heater input and net power transferred
values, and the test section gains and losses.
In experimental work, repeatability of the experiments is one of the main
approaches for the accuracy and reliability of the results. In the experiments of
this thesis, repeatability could not be followed for all of the experiments due to the
fact that the thesis was based on a research project and the schedule was to be
followed strictly in order to finalize the project.
80
3.3 FLOW ORDER
In this section, the two-phase flow experiments are explained in a detailed way by
using a flow order. The sequence of the operation of the devices, the changes
with respect to the measurements are given in Tables 11 and 12.
The first part of the flow order is presented in Table 11 as the sequence from the
computer start-up to refrigerant pump start-up.
Table 11: Flow order part 1: from computer start-up to refrigerant pump start-up
1 Turn on the computer, DC power supply, and the data logger 2 Start the data acquisition program on the computer 3 Check the refrigerant pressure 4 Is it zero? YES Charge the cycle with R-134a
Go back to step 3
NO Continue with step 5 5 Is the refrigerant in liquid phase* **? YES Go to step 7 NO Close the valves of the refrigerant bath
Turn on the bath to decrease the temperature
Continue with step 6 6 Is the liquid refrigerant seen in the sight glass? YES Go to step 7 NO Charge the cycle more with R-134a
Decrease the bath temperature
Go back to step 5 7 Close the flowmeter valves and start the refrigerant pump * Check from the computer ** Check the sight glass
After the pump was started to operate, the flowmeter and the pre-heater were
turned on and so was the water side. Then, by following the flow order presented
in Table 12, the experiments were conducted.
81
Table 12: Flow order part 2: from refrigerant pump start-up until the end
8 Set the refrigerant pump speed to a desired value 9 Is the refrigerant in liquid phase for the whole cycle* **?
YES Continue with step 10 NO Go back to step 6 10 Turn on the flowmeter and open the flowmeter valves 11 Turn on the water bath and pump 12 Turn on the pre-heater and adjust a small power input value 13 Increase the power input of the pre-heater 14 Adjust water temperature w.r.t. R-134a saturation temperature 15 Check the steady-state conditions***
16 Is the water temperature close to and higher than the refrigerant temperature?
YES Continue with step 17 NO Go back to step 14 17 Does the water temperature decrease in the test section? YES Continue with step 18 NO Go back to step 14 18 Is the constant wall-temperature acquired at the test section? YES Continue with step 19 NO Adjust water flow rate Go back to step 14 19 Record the flow rate of the refrigerant (50 readings minimum) 20 Is the refrigerant in two-phase at the inlet* ****? YES Continue with step 21 NO Go back to step 13 21 Note down the time and record the flow rate of water 22 Is the experiment finished for the specified saturation pressure? YES Continue with step 23 NO Go back to step 8 (to change the flow rate) 23 Are the experiments finished for all saturation temperatures? YES END
NO Change the saturation pressure by charging or discharging the refrigerant
Go back to step 8 * Check from the computer ** Check the sight glass *** 50 successful consecutive readings in the range of device accuracy **** Apply the calculations
82
3.4 DATA ANALYSIS
When the 23 step flow order of two-phase flow experiments were finished, the
analyses of the data gathered came into consideration. These analyses can be
classified in two main groups: the analysis of the manual measurements and the
analysis of the recorded measurements by the computer.
The manual measurements were the refrigerant mass flow rate and the water
volumetric flow rate. These measurements were performed at the time of the
experiments since the values were not recorded automatically. For the accuracy
and consistency, they were noted down after the steady-state conditions
achieved and continued at least for 15 minutes. The refrigerant mass flow rate
was recorded from the flowmeter transmitter digital screen in the units of g/s.
Even for the steady-state conditions, oscillations were observed. Therefore, at
least 50 measurements were recorded and the average was taken. The
volumetric flow rate of water was read on the mechanical rotameter manually and
recorded in the experiment data sheet. Since the value did not appear on a digital
screen as in the flowmeter, the value was noted down several times after the
steady-state conditions were reached. The average value of the readings was
accepted as the volumetric flow rate of the water.
The power, temperature, and pressure measurements were the ones that were
recorded in the computer through the data acquisition unit. These measurements
were performed by the power and pressure transducers, RTDs, and
thermocouples. The computer program scanned the data logger to gather the
data for 15 second time intervals during the whole procedure. The read values
were plotted on graphs just after the readings were done. Therefore, the steady-
state was easily followed from the graphs. After the steady-state conditions were
reached, the average value of the power input and the final values of the
pressure of the refrigerant, and the temperatures of the refrigerant at the inlet and
exit of the pre-heater (at the flowmeter exit and the inlet of the test section) were
recorded. These values and the mass flow rate of the refrigerant were used to
calculate the state at the inlet of the test section. When there was a quality value
calculated (two-phase flow was reached), and the other procedure in the test
83
section was followed for the two-phase experiments, the time of the experiment
was noted down and the experiments were continued for different mass flow
rates, saturation pressures, and quality values. The data analysis was left to be
performed when the experiments during that session were completed. The
computer program collected the data in a single file for a session, so it was more
efficient and easier to analyze the data when the session was ended. The
analysis of the data was made in terms of steady-state approach. The
measurements versus the time were drawn as a graph and it was combined with
the times of the mass flow rate record. The averages of 50 readings (minimum) of
the power, temperature, and pressure measurements were taken and they
formed the experiments data. The “Calculations” section continued the procedure
after the experimental data was present.
3.5 CALCULATIONS
The heat transfer coefficient calculation for the two-phase flow of R-134a was the
main aim in the calculations. Through the procedure, many other parameters,
properties, and variables were measured and calculated. The calculations were
performed in two main ways. These are the experimental calculations and the
calculations of correlations in the literature based on the experimental data. The
refrigerant side geometry given in Table 13, was needed to perform the
calculations.
Table 13: The refrigerant side geometry in the test section
Inner diameter (mm) 1.65 Outer diameter (mm) 3.175 Inner perimeter (mm) 5.18
Cross sectional area (mm2) 2.14 Hydraulic diameter (mm) 1.65 Equivalent diameter (mm) 1.65 Test length (mm) 400 Heat transfer area (mm2) 2073.5
84
3.5.1 Experimental Calculations
The calculations of the experiments based on the experimental data were
performed to calculate the average heat transfer coefficient of the refrigerant side
for the varying quality (average quality) values of the refrigerant. As mentioned in
the “Calibrations” section, the refrigerant at the test inlet was desired to be in two-
phase conditions, so the refrigerant was pre-heated before the test section. The
enthalpy of the refrigerant was increased with the pre-heater power input in Eq.
(17) and the determination of the quality of the refrigerant was done by finding
the actual power transfer to the refrigerant by the pre-heater calibration Eq. (18).
The refrigerant pre-heater exit condition was the test section inlet condition,
therefore:
ℎ@6 = ℎ15 (27)
ℎ15 is the test section inlet enthalpy of the refrigerant. Along the test section, the
heat transfer took place from the water side to the refrigerant side. This heat
transfer rate was calculated with Eq. (19). The viscous heating for the water side
was calculated from Eq. (21). When the enthalpy change in the water along the
test section was also added from Eq. (20), the net heat transfer rate could be
found. The exit enthalpy of the test section for the refrigerant could be found with
all other parameters known including the mass flow rate.
ℎ16 = # o + ℎ15 (28)
4 Q is the mass flow rate of the refrigerant, ℎ16 is the test section exit enthalpy of
the refrigerant. Knowing the exit enthalpy and the pressure drop along the test
section, the refrigerant state at the exit could be determined. As mentioned in the
previous sections, the pressure at the exit (inlet pressure - pressure drop) of the
test section was taken into consideration instead of the temperature value. The
temperature of the exit state was found from the pressure and enthalpy values.
The difference of the temperatures at the inlet and exit of the test section
between the calculated and measured values was between -0.69 – 0.17ºC for the
85
inlet and -0.23 – 0.32ºC for the exit. Moreover the quality value at the exit state
was calculated. The temperature at the inlet state was read from the tables with
the same procedure, by combining the pressure (inlet pressure) and the enthalpy.
The wall temperature was needed in the calculations of the average heat transfer
coefficient. The temperature of the wall of the copper tube was measured at 7
locations with 50 mm intervals. The constant-wall temperature approach was
accepted when the maximum temperature difference for these 7 locations was
lower than 0.5°C. Then the average value was taken in the heat transfer
coefficient calculations. The logarithmic mean temperature difference (LMTD)
method was applied to calculate the overall heat transfer coefficient and the
average heat transfer coefficient. The mean temperature difference was not
applied because the temperature of the refrigerant changed along the test
section unlike the wall temperature. The logarithmic mean temperature difference
for the test section is given in Eq. (29).
∆)=> = ,!J!xE/J,!J!x/[*xE
x (29)
)Y is the wall temperature, )15 is the test inlet temperature of the refrigerant, and
)16 is the test exit temperature of the refrigerant. The heat transfer area was
calculated easily from the geometry of the test section. Therefore, the overall
heat transfer coefficient was calculated with Eq. (30).
;5 = # oAE∙∆!m
(30)
The overall heat transfer coefficient, Ui, was based on the inner heat transfer
area, Ai, which is:
5 = 2 ∙ ∙ 5 ∙ (31)
ri is the inner radius of the copper tube and L is the length of the test section. The
overall heat transfer coefficient had the refrigerant convection coefficient and the
86
conduction coefficient across the copper tube. Thus, the average heat transfer
coefficient for the two-phase refrigerant flow is:
ℎ!| = − 1EBC ,1D 1E⁄ /
HJ
(32)
In Eq. (32), ro is the outer radius of the copper tube and kc is the thermal
conductivity of copper. Apart from the heat transfer coefficient calculations, the
pressure drop was recorded along the test section for the two-phase refrigerant
flow. These values are presented in the results section.
3.5.2 Calculations with Correlations in Literature
The heat transfer coefficient calculations were also performed using the
correlations available in the literature. In the flow boiling for internal flow, there
are mainly two types of boiling in terms of the bulk fluid temperature [52]. The
fluid, which is in liquid-phase, is heated along the pipe or channel in which it
flows. The temperature of the fluid increases at the walls of the pipe/channel and
bubble formation occurs. These bubbles collapse when they flow into the bulk
liquid because the temperature of the liquid is lower than the boiling temperature.
This type flow boiling is called sub-cooled boiling. On the other hand, when the
temperature of the liquid also becomes equal to the saturation temperature, two-
phase boiling emerges and the heat transfer coefficient increases by a significant
amount. The two-phase boiling has two forms, which are convective boiling and
nucleate boiling. For low quality values, nucleate boiling dominates heat transfer,
while convective boiling plays the active role for higher quality values. For
moderate values, they both have important effects. The correlations that were
used to calculate the two-phase heat transfer coefficient were based on the
theory mentioned and the outcomes are presented in the Results section to
compare with the experimental results.
87
3.5.2.1 Güngör and Winterton Correlation
Two scientists, Güngör and Winterton [52], studied flow-boiling heat transfer of
halocarbon refrigerants and proposed the correlation given in Eq. (33):
ℎ!| = ℎ[ + 7ℎ_ (33)
ℎ!| is the average heat transfer coefficient for two-phase flow, E is an
enhancement factor, ℎ[ is the heat transfer coefficient for the liquid phase at the
specified condition, S is the suppression factor, and ℎ_ is the pool boiling term.
ℎ[ can be calculated from any single-phase heat transfer coefficient formula
available in the literature with respect to the flow type and Re limitations. The
calculation of E, the enhancement factor, is given in below.
= 1 + 2.4 × 10J3. + 1.37,1 ⁄ /. (34)
In Eq. (34), Bo is the boiling number and 1/Xtt is the Lochart-Martinelli parameter.
Bo is given in Eq. (35) and 1/Xtt is defined in Eq. (36).
= "∙5q
(35)
¡¢¢= £
,J£/.¤ MqM
.¥ ¦ ¦q
. (36)
In Eqs. (35) and (36), q” is the heat flux, G is the mass flux, ilv is the vaporization
enthalpy of the refrigerant, ρ is the density and µ is the viscosity, and l and v are
the subscripts for the liquid and vapor phases.
In Eq. (33), S and ℎ_ are the remaining unknowns. S is defined as:
7 = s1 + 1.15 × 10Jcd[.tJ (37)
88
In this equation, Rel is the Reynolds number for the liquid and it is defined as:
cd[ = ,J£/1E¦q
(38)
The remaining term, ℎ_, is the pool boiling term:
ℎ_ = 55§1.,−¨©§1/J.¥¥ªJ.¥&". (39)
In the final equation of Güngör and Winterton correlation, pr is the reduced
pressure, and M is the molecular weight of the fluid.
3.5.2.2 Chen Correlation
This correlation [52] combines the nucleate boiling, ℎ*0, and convective boiling,
ℎ«0, terms by calculating them separately. The two-phase boiling heat transfer
coefficient is given in Eq. (40).
ℎ!| = ℎ«0 + ℎ*0 = ℎ[%\ + ℎ_7 (40)
ℎ[ and S are the same as in Eq. (33). The boiling enhancement factor, Fo is
calculated from:
%\ = %,1 − ¬/ (41)
% = 1, 1 ⁄ ≤ 0.1 2.35,0.213 + 1 ⁄ /., 1 ⁄ > 0.1¯ (42)
ℎ_ is the pool boiling term and it is calculated by:
ℎ_ = 0.00122 Hq.¤«°,q.±Mq.¤².Ph_ .±
³.±¦q.P¤,5q M /.P (43)
89
kl is the liquid thermal conductivity, cp,l is the liquid specific heat, θb is the wall
superheat, σ is the surface tension, and ∆pv is the pressure for vaporization the
liquid at the specified conditions. ∆pv is calculated from the Clapeyron Equation.
´§µ = ²5q M !lw¢
(44)
3.5.2.3 Bertsch Correlation
Bertsch correlation [53] is similar to the Chen correlation in terms of the separate
calculation of the nucleate boiling and the convective boiling. This correlation is
relatively new and the formulation is given as below.
ℎ!| = ℎ*0,1 − ¬/ + ℎ«0[1 + 80,¬ − ¬/dJ. \] (45)
ℎ*0 is the nucleate boiling heat transfer coefficient and it is calculated from the
same equations of Chen correlation. The confinement number, Co, is:
¶ = ? ³8,MqJM /·nPI ⁄
(46)
The convective boiling term, ℎ«0, is calculated using the following equation.
ℎ«0 = ℎ«\*µ,[,1 − ¬/ + ℎ«\*µ,µ¬ (47)
ℎ«\*µ,[ is the liquid phase convective heat transfer coefficient and ℎ«\*µ,µ is the
vapor phase convective heat transfer coefficient. They are given in the following
equations.
ℎ«\*µ,[ = 3.66 + .∙,·n =⁄ /∙Q6q|1qp.3∙[,·n =⁄ /∙Q6q|1q]P ¸⁄ ∙ Hq
·n (48)
ℎ«\*µ,µ = 3.66 + .∙,·n =⁄ /∙Q6 |1 p.3∙[,·n =⁄ /∙Q6 |1 ]P ¸⁄ ∙ H
·n (49)
90
For the calculations with the correlations, the properties of the refrigerant were
needed. Therefore, for data points of the properties, curve fitting was applied
similar to the water side. The property tables and graphs with curve fittings are
given in Appendix G.
91
CHAPTER 4
EXPERIMENTAL RESULTS AND DISCUSSION
The calculations based on the experimental data were performed in order to find
the two-phase (boiling) heat transfer coefficient of R-134a. Moreover, the
pressure drop of R-134a along the test section was recorded as mentioned
before; therefore, the pressure drop in two-phase flow is presented in addition to
the heat transfer coefficient. In total, 35 experiments were conducted. Due to the
experimental conditions and errors, 7 of them were eliminated. On the other
hand, while presenting the results, one parameter was tried to be fixed and the
other one was investigated in terms of the effect on the heat transfer coefficient
and pressure drop. In such a situation, the successful experimental results were
also selected in terms of the other parameters’ effects, because there were many
variables in the tests. For instance, when the considering effect of saturation
pressure on the heat transfer coefficient for constant quality, the results with
closer mass flux values were selected, because the mass flux had an important
effect the heat transfer coefficient.
The experimental results are presented in two sections: Heat Transfer Coefficient
and Pressure Drop. All data related to the calculations are given for one
experiment as an example in Appendix H. The refrigerant and water side
geometries at the test section, the test conditions, the states of the refrigerant at
the experiment related locations, the properties of both fluids, and the results of
the theoretical and experimental calculations are presented therein.
92
4.1 HEAT TRANSFER COEFFICIENT
The heat transfer coefficient has been calculated in two ways: experimentally and
through the correlations presented above, referred to as theoretically. Since the
constant wall temperature approach was applied in the experiments, the results
were prepared in terms of constant quality and constant mass flux values.
Constant heat flux curves, which are very common in the literature to see the
effect of other variables on the heat transfer coefficient, were not drawn since the
local heat flux along the flow was not uniform.
4.1.1 Constant Quality Results of the Experiments
During the experiments, it was difficult to adjust the quality of the fluid, because
there were many variables that affected the results. When the analyses of the
experimental data were performed and the results were calculated, the
experiments with closer average quality values were selected in order to draw the
graphs. These values were considered as constant quality values. Two graphs
and two tables with the data of these graphs were prepared. In Table 14 and
Figure 27, the saturation pressure effect on the heat transfer coefficient for
constant quality values was examined. The quality values for the constant
assumption are also given in the table. The difficulty of the quality adjustment
caused this assumption, but the results gave an understanding of the effect in a
range of ± 5% for 18% and 25% quality values and ± 10% for 50% quality value.
From Figure 27, it is seen that the heat transfer coefficient does not depend on
the saturation pressure significantly except for 18% quality. Although there are
some oscillations for each and every constant quality value, from this graph, it
can be concluded that the experimental results match most of studies in the
literature claiming the independency of the heat transfer coefficient from
saturation pressure. On the other hand, the heat transfer coefficient value is
larger for lower quality values than higher quality values for the same saturation
pressure value.
93
Table 14: Heat transfer coefficient dependence on saturation pressure for
constant refrigerant quality
18% 25% 50%
Psat hexp x (%) Psat hexp x (%) Psat hexp x (%) bar W/m²K - bar W/m²K - bar W/m²K - 4.96 6324.71 20.42 5.42 4550.70 26.28 6.24 4364.40 45.76 5.25 5890.32 14.94 5.57 5458.17 32.44 6.45 3708.17 47.61
5.25 7024.43 18.63 6.07 4462.30 28.89 6.68 3963.87 60.02 5.55 5750.29 18.20 6.56 4383.87 24.24 6.74 3661.30 61.56 5.66 5404.12 23.32 6.75 3525.54 21.61 7.73 3374.88 52.01
6.89 4764.25 21.36 6.89 4764.25 21.36
7.30 4318.26 30.75
Figure 27: Heat transfer coefficient versus saturation pressure for constant
refrigerant quality
0
1000
2000
3000
4000
5000
6000
7000
8000
4 5 6 7 8
Hea
t tr
ansf
er c
oef
fici
ent
[W/m
2 K]
Saturation pressure [bar]
x = 18%
x = 25%
x = 50%
94
Another analysis about the constant quality results was performed to examine the
effect of mass flux on the heat transfer coefficient. The mass flux range for the
experiments (150 - 450 kg/m2s) was not wide since the experiments were
conducted for low mass flow rates. However, to analyze the effects in the range,
Table 15 and Figure 28 were studied. As in the previous table, the assumed
constant quality values are presented here.
Table 15: Heat transfer coefficient dependence on mass flux for constant
refrigerant quality
18% 25% 50% Mass flux hexp x (%)
Mass flux hexp x (%)
Mass flux hexp x (%)
kg/m²s W/m²K - kg/m²s W/m²K - kg/m²s W/m²K - 158.73 2962.64 23.90 158.73 2962.64 23.90 180.80 3661.30 61.56 219.13 3654.03 14.00 235.33 4462.30 28.89 192.38 3708.17 47.61 246.73 3525.54 21.61 236.40 4318.26 30.75 194.87 4364.40 45.76 255.81 5404.12 23.32 255.81 5404.12 23.32 195.01 3963.87 60.02 325.59 5890.32 14.94 272.38 4764.25 21.36 207.83 5971.38 47.58 339.34 6324.71 20.42 339.34 6324.70 20.42 219.48 5543.35 56.56
343.27 7024.43 18.63
The number of data points for each quality value is not the same because of the
previously mentioned difficulties in the experiment procedure and property, states
and test points.
Figure 28 shows that the heat transfer coefficient increases with the increasing
mass flux values. The trend can be considered as valid for different constant
quality values. Therefore, it is easy to comment that the mass flux and the two-
phase heat transfer coefficient scale with each other in the range of the mass
fluxes studied.
95
Figure 28: Heat transfer coefficient versus mass flux for constant refrigerant
quality
4.1.2 Constant Mass Flux Results of the Experiments
Besides the constant quality results, the constant mass flux results were selected
and analyzed from all of the experimental results. As in the constant quality
analyses, the mass flux constancy could not be adjusted definitely. A change of
the volumetric flow rate provided by the pump was observed due to the material
deterioration (PPS) of the gears of the pump. Achieving two-phase flow,
gathering data successfully, and performing experiments for different test
conditions were focused on during the experiments. For this reason, the ranges
in ∓ 5-10% of the nominal values were accepted as the constant mass flux value.
0
1000
2000
3000
4000
5000
6000
7000
8000
150 200 250 300 350
Hea
t tr
ansf
er c
oef
fici
ent
[W/m
2 K]
Mass flux [kg/m2s]
x = 18%
x = 25%
x = 50%
96
Table 16: Heat transfer coefficient dependence on saturation pressure for
constant refrigerant mass flux
200 250
Psat hexp Mass flux Psat hexp Mass flux bar W/m²K kg/m²s bar W/m²K kg/m²s 6.24 4364.40 194.87 5.42 4550.70 250.05 6.45 3708.17 192.38 5.66 5404.12 255.81 6.65 3654.03 219.13 6.07 4462.30 235.33 6.68 3963.87 195.01 6.56 4383.87 249.64 6.74 3661.30 180.80 7.30 4318.26 236.40
7.18 3118.05 205.27 7.73 3074.88 193.07
Table 16 and Figure 29 reveal the effect of saturation pressure on heat transfer
coefficient for the constant mass flux of the refrigerant in the test section.
Figure 29: Heat transfer coefficient versus saturation pressure for constant
refrigerant mass flux
0
1000
2000
3000
4000
5000
6000
4 5 6 7 8
Hea
t tr
ansf
er c
oef
fici
ent
[W/m
2 K]
Saturation pressure [bar]
G = 200 kg/m2s
G = 250 kg/m2s
97
It can be observed in Table 16 that the mass flux values were between 180 and
220 kg/m2s for the 200 kg/m2s mass flux value and 235 and 260 kg/m2s for the
250 kg/m2s. The saturation pressure effect on the heat transfer coefficient for the
constant mass flux is negligible as in the constant quality analysis. Therefore, the
study of the comparison of the experimental results, the correlations and the
studies in the literature in terms of the saturation pressure effect on the heat
transfer coefficient were not included to this thesis.
The most common type of the analysis of the heat transfer in two-phase flow is
the effect of quality on the heat transfer coefficient for the constant mass flux. The
table and the figure were prepared for two different mass flux values as in the
saturation pressure effect analysis. The quality values calculated were from
11.70% to 67.05% for the mass flux value of 200 kg/m2s and from 12.75% to
29.27% for 250 kg/m2s. The ranges were not for the whole quality values but the
majority of the studies in the literature showed that the heat transfer coefficient
tended to increase with the increasing quality up to 30 – 40 % quality and it
decreased for the higher quality values.
Table 17 and Figure 30 show the experimental results of the effect of quality
variation on the heat transfer coefficient.
The heat transfer coefficient increases with the increasing quality up to 30% for
both mass flux values. The higher quality values are not present for mass flux of
250 kg/m2s, therefore further discussion cannot be performed for that value. For
200 kg/m2s mass flux, the heat transfer coefficient decreases for higher quality
values, so the experimental results are consistent with the studies in the
literature.
98
Table 17: Heat transfer coefficient dependence on quality for constant refrigerant
mass flux
200 250
x (%) hexp Mass flux x (%) hexp Mass flux - W/m²K kg/m²s - W/m²K kg/m²s
14.00 3654.03 219.13 15.15 3335.30 266.20 32.44 5458.17 195.56 21.61 3525.54 246.73 30.95 5426.60 211.85 24.24 4383.87 249.64 42.58 5971.38 207.83 26.28 4550.70 250.05 46.33 5294.66 185.37 28.89 4462.30 235.33 60.02 3963.87 195.01 30.75 4618.26 236.40
61.56 3661.30 180.80 68.50 3118.05 205.27
Figure 30: Heat transfer coefficient versus quality for constant refrigerant mass
flux
0
1000
2000
3000
4000
5000
6000
7000
0 20 40 60 80
Hea
t tr
ansf
er c
oef
fici
ent
[W/m
2 K]
Quality (%)
G = 200 kg/m2s
G = 250 kg/m2s
99
4.1.3 Comparison with Literature
The experimental results are compared with the correlations mentioned in
Chapter 3 and two sets of experimental results presented in Chapter 1;
references [23] (Shiferaw) and [28] (Tibiriçá). The calculations with the
correlations were made with the steps described in the “Calculations with
Correlations in Literature” section. The first comparison was performed for
constant quality. The mass flux was varied from 150 to 350 kg/m2s and the
results are presented in Table 18 and Figure 31 for the experiments and three
correlations.
Table 18: Comparison of experimental results for the heat transfer coefficient for
constant quality and varying mass flux
18% Heat Transfer Coefficient
Mass flux Experimental Güngör Bertsch Chen x (%) kg/m²s W/m²K W/m²K W/m²K W/m²K - 158.73 2962.64 3581.25 3094.91 3061.59 23.90 219.13 3654.03 2877.85 2637.25 1843.13 14.00 255.81 5404.12 5157.85 4702.44 3314.23 23.32 325.59 5890.32 7979.57 4662.81 4575.11 14.94 339.34 6324.71 8652.69 4791.09 4928.33 20.42 343.27 7024.43 8432.82 4760.44 4566.63 18.63
h exp / hcalculated x 100
Mass flux h exp Güngör Bertsch Chen
kg/m²s W/m²K - - - 158.73 2962.64 82.73 95.73 96.77 219.13 3654.03 126.97 138.55 198.25 255.81 5404.12 104.77 114.92 163.06 325.59 5890.32 73.82 126.33 128.75 339.34 6324.71 73.10 132.01 128.33 343.27 7024.43 83.30 147.56 153.82
The constant quality is 18% and the quality values are also presented in Table
18. In order to have a better understanding of the comparison of the experimental
100
results and the results of the previous studies, the ratio of the experimental heat
transfer coefficient over the heat transfer coefficient calculated from the
correlations are also presented in Table 18, as a percentage. It can be seen from
the table that when the extreme results of the Chen correlation are neglected; the
experimental results are in the range of 75% – 150% of the results calculated
from the correlations.
All the results in Figure 31 are similar in terms the mass flux effect on the heat
transfer coefficient, that increasing mass flux leads to an increase in the heat
transfer coefficient. The experimental results are also in a good match with the
correlations. All results are close for low and moderate mass fluxes. The
experimental results are also close to Bertsch and Chen correlation results for
higher values of mass flux values, while Güngör and Winterton correlation gives
higher heat transfer coefficient values.
Figure 31: Experimental results and results of correlations in the literature for
heat transfer coefficient versus mass flux, x = 18%
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
100 150 200 250 300 350 400
Hea
t T
ran
sfer
Co
effi
cien
t [W
/m2 K
]
Mass Flux [kg/m2s]
Experimental
Güngör
Bertsch
Chen
101
The results in terms of the ratios in Table 18 are given in Figure 32. The
experimental results are in a better match with Bertsch correlation results.
Güngör and Winterton correlation gave higher results than the experimental
results, whereas it was the opposite for the Chen correlation.
Figure 32: Heat transfer coefficient ratio versus mass flux, x = 18%
For the rest of all comparisons, the heat transfer coefficient change with quality is
analyzed for a constant mass flux value of 200 kg/m2s. The experimental heat
transfer coefficient was compared with the correlations above and the results of
the experimental studies mentioned in Chapter 1. The comparisons were made
separately in order to comment on the accuracy of the experimental results more
effectively.
The comparison of the heat transfer coefficient with the correlations is presented
in Table 19 and Figure 33.
0
50
100
150
200
250
100 150 200 250 300 350 400
hex
p/ h
calc
ula
ted
x 1
00
Mass Flux [kg/m2s]
Güngör
Bertsch
Chen
102
Table 19: Comparison of experimental results for the heat transfer coefficient for
varying quality
x (%) Mass flux Experimental Güngör Bertsch Chen - kg/m²s W/m²K W/m²K W/m²K W/m²K
14.00 219.13 3654.03 2877.85 2637.25 1843.13 30.95 211.85 5426.60 4756.95 4261.66 2711.90 42.58 207.83 5971.38 4578.98 4200.61 2358.72 46.33 185.37 5294.66 5005.38 4364.83 2828.96 60.02 195.01 3963.87 3924.13 3780.37 2165.83 61.56 180.80 3661.30 4040.46 3705.83 2319.03 68.50 205.27 3118.05 4045.40 3662.16 2299.16
It can be observed in Fig. 33 that the experimental results, Güngör and Winterton
correlation, and Bertsch correlation show a maximum deviation at about 40%
quality and tend to decrease for higher quality values. Bertsch correlation gives
lower heat transfer coefficient values.
Figure 33: Experimental results and results of correlations in the literature for the
heat transfer coefficient versus quality
1000
2000
3000
4000
5000
6000
0 20 40 60 80
Hea
t T
ran
sfer
Co
effi
cien
t [W
/m2 K
]
Quality (%)
Experimental
Güngör
Bertsch
Chen
103
Table 20 and Figure 34 are the comparison of the experimental results with the
experimental results of Tibiriçá [28]. This study was performed with constant heat
flux approach and different heat fluxes were applied. The authors gave the
results of the study of Yan [28] and it is also adapted.
Table 20: Comparison of experimental results with Tibiriçá [28] in terms of quality
dependence
G = 200 kg/m2s Yan [28]
x (%) Experimental x (%) (q" = 5 kW/m2) - W/m²K - W/m²K
14.00 3654.03 10 2950 32.44 5458.17 16.5 2700 30.95 5426.60 20 2700 42.58 5971.38 29.5 3100 46.33 5294.66 40 3050 60.02 3963.87 50 3450 61.56 3661.30 54 3500 68.50 3118.05 70 3600
83 3800
Yan [28] hexp [28]
x (%) (q" = 15 kW/m2) x (%) (q" = 5 kW/m2) - W/m²K - W/m²K
43 3650 15 2900 54 3950 25 2950 64 4200 35 2650 75 4700 45 3000 84 5500 55 3450 94 6600 65 3800
75 4300 85 4800 95 5800
For the increasing quality values, the heat transfer coefficient increases for the
compared results. There is no peak point for the quality value, which is contrary
to the majority of the studies, but also matches some of the studies in the
104
literature. When the results are compared with the experimental results, it is
observed that the experimental heat transfer coefficient is close to the results in
literature. The tendency difference is noted down in this figure.
Figure 34: Heat transfer coefficient versus quality for current experiments and
results of Tibiriçá [28]
The comparison for the constant mass flux and varying quality was also
performed for another experimental study by Shiferaw [23]. Similar to the
previous compared study, the author applied different values of constant heat flux
for separate sets of experiments and collected the data to find the two-phase
heat transfer coefficient.
The results of the study and the experimental results are presented in Table 21
and Figure 35. The heat flux was varied from 17 to 69 kW/m2K. These values are
relatively higher than the experimental conditions. The heat flux values of the
experiments were not presented, but the average heat flux along the test section
0
1000
2000
3000
4000
5000
6000
7000
0 20 40 60 80 100
Hea
t T
ran
sfer
Co
effi
cien
t [W
/m2 K
]
Quality (%)
Experimental
Yan [28] (q" = 5 kW/m2)
Yan [28] (q" = 15 kW/m2)
hexp [28] (q" = 5 kW/m2)
105
for the refrigerant side was in the range of 4 – 20 kW/m2K. Therefore the
experimental results were expected to deviate from the results of the study but
the chance for the investigation of the tendency would be a good opportunity to
evaluate the experimental set-up.
Table 21: Comparison of experimental results with Shiferaw [23] in terms of
quality dependence
G = 200 kg/m2s hexp [23] hexp [23]
x (%) Experimental x
(%) (q" = 17 kW/m2) x
(%) (q" = 27 kW/m2) - W/m²K - W/m²K - W/m²K
14.00 3654.03 1 6050 6 9600 32.44 5458.17 2 6000 13 9600 30.95 5426.60 4 5950 20 10000 42.58 5971.38 6 5900 27 11000 46.33 5294.66 7 5800 31 11700 60.02 3963.87 9 5850 34 11400 61.56 3661.30 12 5900 68.50 3118.05 14 6200
16 6800 18 6100
hexp [23] hexp [23] hexp [23]
x (%) (q" = 34 kW/m2) x
(%) (q" = 53 kW/m2) x
(%) (q" = 69 kW/m2) - W/m²K - W/m²K - W/m²K 2 9000 2 14000 7 14200 5 9250 7 15300 15 15300 9 9000 13 15300 23 15500 13 8900 19 15300 31 15700 16 8850 25 15100 38 15500 20 8750 31 14900 47 15300 25 8800 37 14800 54 14100 28 9000 43 15000 62 11000 31 9050 49 14100 69 8000 36 9950 56 14100 77 6000 39 10100 62 11200 85 4200 42 9050 67 8000 93 3200
106
Figure 35 shows that the heat transfer coefficient increases for the moderate heat
flux experiments, but the same comment cannot be made for the high quality
results. For the high heat flux experiments, heat transfer coefficient does not
change significantly up to about 50% quality and then there was a dramatic
decrease for the high quality values, which could be caused by dry-out as it was
explained in many studies in the literature.
Figure 35: Heat transfer coefficient versus quality for the current experiments and
those of Shiferaw [23], G = 200 kg/m2s
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
0 20 40 60 80 100
Hea
t T
ran
sfer
Co
effi
cien
t [W
/m2 K
]
Quality (%)
Experimental
hexp [23] (q" = 17 kW/m2)
hexp [23] (q" = 27 kW/m2)
hexp [23] (q" = 34 kW/m2)
hexp [23] (q" = 53 kW/m2)
hexp [23] (q" = 69 kW/m2)
107
4.2 PRESSURE DROP
The pressure drop along the test section was measure using the DPT. The
results are presented in this section. No calculations were performed but the
effects of different parameters on the pressure drop were observed by preparing
tables and figures. As for the heat transfer coefficient, the constant quality and
the constant mass flux effects on the pressure drop were examined.
The pressure drop was measured in the units of [bar], and it was not changed in
the tables and figure. Table 22 and Figure 36 present the effect of mass flux on
pressure drop. The constant assumed quality values are also presented.
Table 22: Dependence of pressure drop on mass flux for constant quality
18% 25% 50% Mass flux ∆P x (%)
Mass flux ∆P x (%)
Mass flux ∆P x (%)
kg/m²s bar - kg/m²s bar - kg/m²s bar -
219.13 0.02 14.00 195.56 0.03 32.44 171.51 0.05 58.61
246.73 0.03 21.61 211.85 0.03 30.95 180.80 0.06 61.56 266.20 0.05 15.15 236.40 0.05 30.75 185.37 0.05 46.33 272.38 0.05 21.36 241.41 0.05 30.63 193.07 0.05 52.01 325.59 0.09 14.94 249.64 0.04 24.24 194.87 0.05 45.76 339.34 0.15 20.42 252.83 0.06 27.11 195.01 0.06 60.02 343.27 0.16 18.63 255.81 0.06 23.32 199.03 0.07 56.33
207.83 0.06 47.58 219.48 0.07 56.56
The pressure drop increases with the increasing mass flux values for all three
qualities. The increase for the 18% quality is much more dramatic for the
available data points than it is for the other values.
108
Figure 36: Pressure drop versus mass flux for constant refrigerant quality
The pressure drop change with quality change for constant mass flux was the
second investigation subject about the pressure drop considerations. The
pressure drop was considered as an important factor in two-phase flow due to the
fact that it is larger than the single-phase (liquid) pressure drop. Table 23 and
Figure 37 give the values of pressure drop with changing quality. The constant
assumed mass flux values are presented in the table for the possible
experimental errors. The graph shows the increase of the pressure drop with the
increasing quality values. It is an expected result when it is compared with the
studies in the literature.
0,00
0,04
0,08
0,12
0,16
0,20
150 200 250 300 350 400
Pre
ssu
reD
rop
[bar
]
Mass flux [kg/m²s]
x = 18%
x = 25%
x = 50%
109
Table 23: Dependence of pressure drop on quality for constant refrigerant mass
flux
200 x (%) ∆P Mass flux
- bar kg/m²s 14.00 0.02 219.13 32.44 0.03 195.56 30.95 0.03 211.85 45.76 0.05 194.87 46.33 0.05 185.37 47.61 0.05 192.38 52.01 0.05 193.07 60.02 0.06 195.01 61.56 0.06 180.80 68.50 0.06 205.27
Figure 37: Pressure drop versus quality for the constant refrigerant mass flux,
G = 200 kg/m2s
0,00
0,01
0,02
0,03
0,04
0,05
0,06
0,07
0 20 40 60 80
Pre
ssu
reD
rop
[bar
]
Quality (%)
110
CHAPTER 5
CONCLUSIONS
5.1 SUMMARY & CONCLUSIONS
In this thesis, the process and the progress of the experimental set-up
construction to perform experiments of two-phase R-134a flow in a 1.65 mm tube
were presented. The advantages of heat transfer in small scale media were
explained and the need for the low flow rate and constant wall temperature
approach studies in the literature were pointed out. Many studies in the literature
were given with their relations to the current study. Based on these studies and
the needs in the establishment of the infrastructure about the subject, the
motivation and the aims of the experiments were shaped.
From the design to the experiments, the set-up was described in detail in terms of
the devices used, the placement and operation of the devices, the experimental
conditions, the experimental procedures, and the experimental results. These
results were compared with some of the available theoretical and experimental
studies in the literature and the comments and discussions were made.
This study was focused on the set-up construction, and the approaches and data
analyses for the two-phase flow experiments. The results were not enough due to
the mentioned difficulties in terms of sufficient data points to propose any
correlations. On the other hand, the available results were observed to match
with the results of the compared correlations and experimental studies. The
results of both heat transfer coefficient calculations and pressure drop did not
111
deviate much; in addition, the tendencies were compatible with the majority of the
studies in the literature.
To sum up, the study aimed to support the widening of the two-phase flow
database in small scale media experimentally. The more the studies about this
subject with constant wall temperature approach; the higher chance is the use in
domestic refrigerators.
5.2 SUGGESTIONS FOR FUTURE WORK
The experimental set-up is ready to operate for other experiments and to test
many other tubes and channels. By only changing the test section, the set-up can
be used with different tubes and channels. For the future studies with the
experimental set-up, proper vacuuming after each opening of the cycles is
strongly recommended. Moreover, an extra pump parallel to the existing pump
can be mounted on the system to prevent damage in the gears of the pump since
they tend to wear out with the flow of the refrigerant in vapor phase. The
calibrations for the pre-heater and the test section are to be performed for any
modifications of the test section in terms of major changes in flow geometry.
In the existing study, the focus was on the two-phase flow. Single-phase flow of
the refrigerant can be also studied for not only laminar, but also turbulent flow. On
the other hand, the constant wall temperature approach with single-phase flow
would be more difficult to apply.
Besides from the different geometry tests and single-phase flow, the working
fluids can be changed and the experiments for other refrigerants and other fluids
can be made. Since there are two independent cycles in the set-up, with the
consideration of the compatibility of the devices, both sides single-phase, both
sides two-phase, and one-side liquid one-side gas flow experiments can be
performed. Moreover, numerical studies could be performed in addition to the
experimental studies and the results could be compared to verify the numerical
model. With these possible future work subjects, many researches can be made
and valuable contributions to the literature about the internal flow can be added.
112
REFERENCES
[1] Energy Information Administration, Electricity Consumption by End Use in U.S. Households, 2001, www.eia.doe.gov, last visited on 31.01.2011.
[2] Advanced Technology Program, Mechanical Refrigeration, www.atp.nist.gov, last visited on 31.01.2011.
[3] Kandlikar, S.G., Heat Transfer Mechanisms During Flow Boiling in Microchannels, ASME, 2004, 126, pp. 8-16.
[4] Tuckerman, D.B. and Pease, R.F., Optimized Convective Cooling using Micromachined Structure, Journal of the Electrochemical Society, 1982, 129(3), pp. C98.
[5] Choi, S.B., Barron R.F., and Warrington R.O., Fluid Flow and Heat Transfer in Microtubes, Micromechanical Sensors, Actuators, and Systems, ASME DSC, 1991, 32, pp. 123-134.
[6] Harms, T.M., Kazmierczak, M.J., and Gerner, F.M., Developing Convective Heat Transfer in Deep Rectangular Microchannels, International Journal of Heat Fluid Flow, 1999, 20, pp. 149-157.
[7] Jung, J.Y., Kwak, H.Y., Fluid flow and heat transfer in microchannels with rectangular cross section, Heat Mass Transfer, 2007, 44, pp. 1041-1049.
[8] Park, H.S., Punch, J., Friction factor and heat transfer in multiple microchannels with uniform flow distribution, Int. J. Heat and Mass Transfer, 2008, 51, pp. 4535-4543.
[9] Kandlikar, S.G., Fundamental issues related to flow boiling in minichannels and microchannels, Experimental Thermal and Fluid Science, 2002, 26, pp. 389-407.
[10] Qu, W., Mudawar, I., Flow boiling heat transfer in two-phase micro-channel heat sinks – I. Experimental investigation and assessment of correlation methods, Int. J. Heat and Mass Transfer, 2003, 46, pp. 2755-2771.
[11] Saitoh, S., Daiguji, H., Hihara, E., Effect of tube diameter on boiling heat transfer of R-134a in horizontal small-diameter tubes, Int. J. Heat and Mass Transfer, 2005, 48, pp. 4973-4984.
[12] Bertsch, S.S., Groll, E.A., Garimella, S.V., Effects of heat flux, mass flux, vapor quality, and saturation temperature on flow boiling heat transfer in microchannels, Int. J. Multiphase Flow, November 2008, 35, pp. 142-154.
113
[13] Revellin, R., and Thome, J.R., Experimental Investigation of R-134a and R-245fa Two-Phase Flow in Microchannels for Different Flow Conditions, International Journal of Heat and Fluid Flow, 2007, 28, pp. 63-71.
[14] Bertsch, S.S., Groll, E.A., Garimella, S.V., Refrigerant flow boiling heat transfer in parallel microchannels as a function of local vapor quality, Int. J. Heat and Mass Transfer, 2008, 51, pp. 4775-4787.
[15] Yan, Y.Y., Lin, T.F., Evaporation heat transfer and pressure drop of refrigerant R-134a in a small-pipe, Int. J. Heat and Mass Transfer, 1998, 41, pp. 4183-4194.
[16] Liu, J.T., Peng, X.F., Yan, W.M., Numerical study of fluid flow and heat transfer in microchannel cooling passages, Int. J. Heat and Mass Transfer, 2006, 50, pp. 1855-1864.
[17] Liu, D., Garimella, S.V., Flow Boiling Heat Transfer in Microchannels, ASME Journal of Heat Transfer, 2008, 129, pp. 1321-1332.
[18] Agostini, B., Bontemps, A., Vertical flow boiling of refrigerant R134a in small channels, Int. J. Heat and Fluid Flow, 2005, 26, pp. 296-306.
[19] Djordjevic, E., Kabelac, S., Flow boiling of R134a and ammonia in a plate heat exchanger, Int. J. Heat and Mass Transfer, 2006, 31, pp. 1-19.
[20] De Rossi, F., Mauro, A.W., Rosato, A., Local heat transfer coefficients and pressure gradients for R-134a during flow boiling at temperatures between -9ºC and +20ºC, Energy Conversion and Management, 2009, 50, pp. 1714-1721.
[21] Ong, C.L., Thome, J.R., Flow boiling heat transfer of R134a, R236fa and R245fa in a horizontal 1.030 mm circular channel, Experimental Thermal and Fluid Science, 2009, 33, pp. 651-663.
[22] Choi, K.I., Pamitran, A.S., Oh, C.Y., Oh, J.T., Boiling heat transfer of R-22, R-134a, and CO2 in horizontal smooth minichannels, Int. J. Refrigeration, 2007, 30, pp. 1336-1346.
[23] Shiferaw, D., Karayiannis, T.G., Kenning, D.B.R., Flow boiling in a 1.1 mm tube with R134a: Experimental results and comparison with model, Int. J. Thermal Sciences, 2009, 48, pp. 331-341.
[24] Matkovic, M., Cavallini, A., Col, D.D., Rossetto, L., Experimental study on condensation heat transfer inside a single circular minichannel, Int. J. Heat and Mass Transfer, 2009, 52, pp. 2311-2323.
[25] Revellin, R., Mishima, K., Thome, J.R., Status of prediction methods for critical heat fluxes in mini and microchannels, Int. J. Heat and Fluid Flow, 2009, 30, pp. 983-992.
[26] In, S., Jeong, S., Flow boiling heat transfer characteristics of R123 and R134a in a micro-channel, Int. J. Multiphase Flow, 2009, 35, pp. 987-1000.
114
[27] Park, J.E., Thome J.R., Critical heat flux in multi-microchannel copper elements with low pressure refrigerants, Int. J. Heat and Mass Transfer, 2010, 53, pp. 110-122.
[28] Tibiriçá, C. B., Ribatski, G., Flow boiling heat transfer of R134a and R245fa in a 2.3 mm tube, Int. J. Heat and Mass Transfer, 2010, doi:10.1016/j.ijheatmasstransfer.2010.01.038
[29] Raja, B., Mohan Lal, D., Saravanan, R., Flow boiling heat transfer study of R-134a/R-290/R-600a mixture in 9.52 and 12.7 mm smooth horizontal tubes: Experimental Investigation, Experimental Thermal and Fluid Science, 2009, 33, pp. 542-550.
[30] Ding, G., Hu Haitao, D., Huang, X., Deng, B., Gao, Y., Experimental investigation and correlation of two-phase frictional pressure drop of R410A–oil mixture flow boiling in a 5mm microfin tube, International Journal of Refrigeration, 2009, 32, pp. 150-161.
[31] Chen, C. A., Lee, C. Y., Lin, T. F., Experimental study of R-134a evaporation heat transfer in a narrow annular duct, Int. J. Heat and Mass Transfer, 2010, 53, pp. 2218-2228.
[32] Saisorn, S., Kaew-on, J., Wongwises, S., Flow pattern and heat transfer characteristics of R-134a refrigerant during flow boiling in a horizontal circular mini-channel, Int. J. Heat and Mass Transfer, 2010, 53, pp. 4023-4038.
[33] Consolini, L., Thome, J. R., Micro-channel flow boiling heat transfer of R-134a, R-236fa, and R-245fa, Microfluid Nanofluid, 2009, 6, pp. 731-746.
[34] Kakaç, S., Yener, Y., Convective Heat Transfer, Middle East Technical University, Ankara, 1980.
[35] Cole Parmer, Thermo Scientific NESLAB RTE-series Refrigerated Circulating Baths, http://www.coleparmer.com/catalog/product_view.asp?sku=1350030&pfx=KH, last visited on 31.01.2011.
[36] Cole Parmer, Benchtop Digital Drive, http://www.coleparmer.com/catalog/Product_view.asp?sku=7401455&pfx=KH&referred_id=2269, last visited on 31.01.2011.
[37] Rhenoik, RHM Coriolis Flow Sensors & RHE Coriolis Flow Transmitters, http://www.ge-mcs.com/, last visited on 31.01.2011.
[38] Cole Parmer, Glass-tubing, http://www.coleparmer.com/catalog/Product_view.asp?sku=3474200&pfx=KH&referred_id=2269, last visited on 31.01.2011.
[39] Cole Parmer, Compact Pressure Transducer, http://www.coleparmer.com/catalog/product_view.asp?sku=6834740, last visited on 31.01.2011.
115
[40] Cole Parmer, Oakton Integral handle RTD probes, http://www.coleparmer.com/catalog/Product_view.asp?sku=0811780&pfx=KH&referred_id=2269, last visited on 31.01.2011.
[41] Cole Parmer, 30-gauge wire, http://www.coleparmer.co.uk/catalog/product_view.asp?sku=0854204&pfx=KH, last visited on 31.01.2011.
[42] Cole Parmer, 303 SS Flowmeters for Water, http://www.coleparmer.com/catalog/product_view.asp?sku=3220502, last visited on 31.01.2011.
[43] Cole Parmer, Cole Parmer® Wet/Wet Differential Pressure Transmitters, http://www.coleparmer.com/catalog/Product_view.asp?sku=6807156&pfx=KH&referred_id=2269, last visited on 31.01.2011.
[44] Egerate, MCH TECHNIC 305D-2 DC Power Supply, http://www.egerate-store.com/Default.aspx?_Args=ProductInfo,43&LNG=EN last visited on 31.01.2011.
[45] Agilent, 34970A Data Acquisition / Data Logger Switch Unit, http://www.home.agilent.com/agilent/product.jspx?cc=TR&lc=eng&ckey=1000001313:epsg:pro&nid=-33640.536881544.00&id=1000001313:epsg:pro, last visited on 31.01.2011.
[46] Thermo Electric, CNN-11300-TX Standard Connector Plug, www.thermo-electric.nl, last visited on 31.01.2011.
[47] Engineering Toolbox, Ethylene Glycol Heat-Transfer Fluid, http://www.engineeringtoolbox.com/ethylene-glycol-d_146.html, last visited on 31.01.2011.
[48] Kerpiççi, H., Personal communication.
[49] Industrial Insulation, Armaflex® Pipe Insulation, www.industrialinsulation.com, last visited on 31.01.2011.
[50] Morini, G., L., Viscous heating in liquid flows in micro-channels, Int. J. Heat and Mass Transfer, 2005, 48, pp. 3637-3647.
[51] Bejan, A., Advanced Engineering Thermodynamics, Wiley, New York, p. 239-263, 1988.
[52] Kakaç, S. and Liu, H., Heat Exchangers: Selection, Rating, and Thermal Design, CRC, Florida, 2002.
[53] Bertsch, S.S., Groll, E.A., Garimella, S.V., A composite heat transfer correlation for saturated flow boiling in small channels, International Journal of Heat and Mass Transfer, 2009, 52, pp. 2110-2118.
116
[54] Sonntag, R., E., Borgnakke, C., Van Wylen, G., J., Fundamentals of Thermodynamics, Wiley, 6th edition, Phoenix, 2003.
117
APPENDIX A
SPECIFICATIONS OF DEVICES IN THE EXPERIMENTAL SET-UP
A.1 REFRIGERATED CIRCULATING BATH
This bath is used for both cycles to condition the fluids inside [35].
Brand Name: Cole Parmer
Catalog Number: KH-13500-30
Bath Capacity 7 Liters
Temperature Range -40 to 200°C
Temperature stability ±0.01°C
Temperature control PID
Temperature setting Digital
Temperature display LED
Temperature sensor 100 Ω Pt RTD
Cooling capacity 800 W @ 20°C, 650 W @ 0°C, 500 W @ -10°C
Wattage: 2550 W total, 2000 W heater
Pressure pump: Max flow 15 L/min, Max head 5 psi
Compressor 1/2 hp
Refrigerant R-404A
Bath opening 6 5/8" x 7 1/4”
Working depth 6"
Overall dimensions 26 5/8"W x 11 3/8"H x 18 7/8"D
Power Input 230 VAC, 50 Hz
Amperes 12 A
118
A.2 DIGITAL GEAR PUMP SYSTEM
This pump is used to circulate the refrigerant and to adjust the flow rate of the
refrigerant [36].
Brand Name: Cole Parmer
Catalog Number: KH-74014-55
Wetted parts Body: 316 SS
Gears: PPS
Seals: PTFE
Flow rate: 0.316 mL/rev
18.96 mL/min at 60 rpm
1137.6 mL/min at 3600 rpm
Differential pressure 75 psi (max)
Max system pressure 300 psi
Max temperature 40°C (system)
Temperature range -46 to 54°C
Port size 1/8" NPT(F)
Viscosity 0.2 to 1500 cp
Dimensions 7 3/4"L x 11 1/2"W x 7 1/4"H
Power VAC 220 VAC
119
A.3 MICRO-FLOWMETER
This micro-flowmeter is used to measure the mass flow rate. It is Coriolis type
mass flowmeter [37].
Brand Name: Rheonik
Flowmeter Model: RHM 04 Universal Coriolis Mass Flowmeter with particular
Fast Response
Pressure up to 250 bar
Range from 0.1 kg/min to 10 kg/min
Minimal flow 0.05 kg/min
Response time 30 ms and better
Flow Accuracy 0.10%
Repeatability 0.05%
Transmitter Model: RHE 08 Rheonik advanced transmitters
Version Wall mounting
Analog Outputs Two (0/4 - 20/22 mA)
Pulse/frequency 1 output - 2 inputs
Supply voltages Available in all common
Multifunctions density, brix, concentration
Power consumption < 15 W
Temperature range -40°C to 60°C
120
A.4 GLASS TUBE
The glass tube was selected to be used in the test section in the first design [38],
but was changed due to the reasons and complications mentioned in the
“Preparation” and “Modifications” parts.
Brand Name: Cole Parmer
Catalog Number: KH-34742-00
Outside Diamter: 3.0 mm
Thickness: 0.6 mm
Material: Borosilicate Glass
Softening point: 821ºC
Working point: 1252 ºC
Length: 48” (each piece)
A.5 COMPACT PRESSURE TRANSDUCER
Compact pressure transducers are used for both cycles to determine the
pressure of the system [39].
Brand Name: Cole Parmer
Catalog Number: KH-68347-40
Output 4-20 mA
Process connection 1/4" NPT (M)
Range 200 psia
Accuracy 0.5% full scale
Power 8 to 30 VDC
121
A.6 RTD PROBE
RTDs are used in both cycles to determine the temperature at the specified
locations from the fluids flowing inside the tubes [40].
Brand Name: Cole Parmer
Catalog Number: KH-08117-80
Temp range -50 to 500°C
Probe length 2"
Diameter 0.093"
Time constant 10 seconds
Sheath material 316 SS
A.7 THERMOCOUPLE WIRE
Thermocouples are used to measure the temperature at the walls of the materials
in which fluids flow [41].
Brand Name: Cole Parmer
Catalog Number: KH-08542-04
Length 1000 ft
Temp range -200 TO 204°C
Error limit ±1.5% reading from -200 to -65°C;
±1°C from -65 to 130°C
±0.75% reading from 130 to 350°C
Type T
Advantages Chemical, moisture and abrasion resistance
Gauge 30
ID 0.0100" (0.25 mm) dia
OD 0.030" x 0.048"
Model T30-2-506
122
A.8 ROTAMETER
Rotameters are used to measure volumetric flow rate of water at two locations
along the water cycle [42].
Brand Name: Cole Parmer
Catalog Number: EW-32205-02 (KH-32205-02)
Media water
Max operating temperature 240°F (115°C)
Max pressure 6000 psi
Flow rate water 0.2 to 1.8 LPM;
Accuracy ±2% full-scale
Connections 1/4" NPT(F)
Repeatability ±1.0%
Housing Material 304 Stainless Steel
Dimensions 1 3/4” W x 4 13/16” H x 1 15/16” D
A.9 DIFFERENTIAL PRESSURE TRANSDUCER
Differential pressure transducer reads the pressure drop at the test section for the
refrigerant flow [43].
Brand Name: Cole Parmer
Catalog Number: KH-68071-56
Dimensions 3"L x 2 1/8"W x 3"H
Output 4 to 20 mA
Operating temperature 0 to 175°F (-17 to 80°C)
Process connection 1/4" NPT(F)
Range 0 to 5.0 psid
Accuracy ±0.25% full scale
Power 11 to 30 VDC (unregulated)
123
A.10 DC POWER SUPPLY
DC power supply is used to power the pressure transducers [44].
Brand Name: MCH Technic
Catalog Number: MCH 305D-2 DC Power Supply
Voltage: 0-30V Adjustable
Current: 0-5A Adjustable
Dual-Output DC Power Source Two Independent: 30V-5A
With Serial Button: 60V-5A
With Parallel Button: 30V-10A
Voltage 0 to Descripted Continuously Variable
Stability Voltage: ≤0.01%+2mV
Load: ≤0.0%+2mV
Recover Time ≤100µS
Ripple & Noise ≤1MVRMS(efficent)
Temperature Factor ≤300PPM/ºC
Current 0 to Descripted Continuously Variable
Current, Load Stability≤0.2%+3mA
Ripple & Noise ≤3mArms
Working Condition
Line Power Supply 220V±10% 50/60Hz
Operating Temperature: -10ºC to 40ºC Moisture≤90%
Storage Temperature: -20ºC to 80ºC Moisture≤80%
124
A.11 DATA ACQUISITION SYSTEM
Data acquisition system is used to gather the pressure and temperature readings
from pressure transducers, RTDs, and thermocouples and transfer them to the
computer interface [45].
Brand Name: Agilent
Catalog Number: 34970A Data Acquisition / Switch Unit
Slot 3
Interface GPIB and RS232
Multimeter Digital and internal with 6 1/2-digit (22-bit)
Scanning up to 250 channels per second
Modules 8 switch and control plug-in
Memory 50k readings
Alarm Hi/Lo on each channel
Measurement Thermocouples, RTDs and thermistors, ac/dc volts and current;
resistance; frequency and period
Software Free Bench Link data logger software enables tests without
programming
Brand Name: Agilent
Catalog Number: 34901A 20-Channel Multiplexer + 2 Current Channels Module
(Built-in thermocouple reference junction)
Model description 34901A 20 ch Multiplexer+ 2 current channels
Type 2-wire armature (4-wire selectable)
Speed(ch/sec) 60
Max volts 300 V
Max amps 1 A
125
APPENDIX B
SKETCH OF EXPERIMENTAL SET-UP IN THE PRELIMINARY DESIGN
Figure B.1: Sketch of the experimental set-up in the preliminary design
126
APPENDIX C
CONNECTIONS OF THE EXPERIMENTAL SET-UP IN THE PRELIMINARY
DESIGN
Table C.1: The necessary pipes for the set-up connections
Pipe (diameter) Length (mm)
1/8" pipe 350
1/4" pipe 5800
1/4" tube 2000
Table C.2: Elbows and T-connections
Elbow and T-Connections #
1/4" NPT (M) pipe (1000 mm) 1
1/4" NPT (F) * 1/4" tube elbow 4
1/4" NPT (F) * 1/4" NPT (F) elbow 7
(2) 1/4" NPT (F) * (1) 1/4" NPT (M) T-connection 1
(3) 1/4" NPT (F) T-connection 3
(2) 1/4" NPT (F) * (1) 1/8" NPT (F) T-connection 1
(1) 1/4" NPT (F) * (1) 1/4" NPT (M) * (1) 1/8" NPT (F) T-connection 1
127
Table C.3: Refrigerant (R-134a) cycle connectors and pipes
Location Material
T1_3 1/4" NPT (M) pipe (300 mm)
3 1/4" NPT (F) * 1/4" tube elbow
3_bath 1/4" tube (300 mm)
4 1/4" tube (300 mm)
4_5 1/4" tube (300 mm)
5 1/4" NPT (F) * 1/4" tube elbow
5_6 1/4" NPT (M) pipe (300 mm)
6 1/4" NPT (M) pipe (300 mm)
6_T2 1/4" NPT (M) pipe (100 mm)
T2 (2) 1/4" NPT (F) * (1) 1/4" NPT (M) T-connection
T2-P 1/4" NPT (F) pipe (100 mm)
T2_1 1/4" NPT (M) pipe (100 mm)
1 1/4" NPT (F) * 1/4" NPT (F) elbow
pipe (1_2) 1/4" NPT (M) pipe (1000 mm)
elbow (1_2) 1/4" NPT (F) * 1/4" NPT (F) elbow
elbow_2 1/4" NPT (M) pipe (300 mm)
T3 (3) 1/4" NPT (F) T-connection
T3_elbow (DPT inlet) 1/4" NPT (M) pipe (100 mm)
elbow (DPT inlet) 1/4" NPT (F) * 1/4" NPT (F) elbow
DPT inlet_DPT 1/4" NPT (M) pipe (300 mm)
DPT_DPT exit 1/4" NPT (M) pipe (300 mm)
elbow (DPT exit) 1/4" NPT (F) * 1/4" NPT (F) elbow
elbow_T1 1/4" NPT (M) pipe (300 mm)
T1 (3) 1/4" NPT (F) T-bağlantısı
test_T1 1/4" NPT (M) pipe (100 mm)
T2_test 1/4" NPT (M) pipe (100 mm)
128
Table C.4: Water cycle connectors and pipes
Location Material
test_7 1/4" NPT (M) pipe (100 mm)
7 1/4" NPT (F) * 1/4" NPT (F) elbow
7_9 pipe 1/4" NPT (M) pipe (200 mm)
7_9 elbow 1/4" NPT (F) * 1/4" NPT (F) elbow
7_9 pipe2 1/4" NPT (M) pipe (200 mm)
9 (T4) (2) 1/4" NPT (F) * (1) 1/8" NPT (F) T-connection
9_8 1/4" NPT (M) pipe (300 mm)
8 1/4" NPT (F) * 1/4" tube elbow
8_bath 1/4" tube (300 mm)
bath_12 1/4" tube (300 mm)
12 1/4" NPT (F) * 1/4" tube elbow
12_rotameter 1/4" NPT (M) pipe (300 mm)
rotameter_10 1/4" NPT (M) pipe (300 mm)
9_pump 1/8" NPT (M) pipe (300 mm)
pump_T5 1/8" NPT (M) pipe (100 mm)
T5 (2) 1/4" NPT (F) * (1) 1/8" NPT (F) T-connection
T5_P 1/4" NPT (M) pipe (100 mm)
T5_10 (T6) 1/4" NPT (M) pipe (100 mm)
10 (3) 1/4" NPT (F) T-connection
10_elbow 1/4" NPT (M) pipe (300 mm)
Elbow 1/4" NPT (F) * 1/4" NPT (F) elbow
elbow_rotameter 1/4" NPT (M) pipe (100 mm)
rotameter_11 1/4" NPT (M) pipe (100 mm)
11_test 1/4" NPT (M) pipe (100 mm)
129
APPENDIX D
TECHNICAL DRAWINGS OF THE PARTS IN THE TEST SECTION
D.1 TECHNICAL DRAWING OF THE FLANGE
Figure D.1: Technical drawing of the flange
130
D.2 TECHNICAL DRAWING OF THE SHELL
Figure D.2: Technical drawing of the shell
131
D.3 TECHNICAL DRAWINGS OF THE RING AND THE BUMPER
Figure D.3: Technical drawings of the ring and the bumper
132
APPENDIX E
THE CALIBRATION TABLES AND GRAPHS FOR MEASUREMENT DEVICES
E.1 A SAMPLE CALIBRATION CURVE FOR A CALIBRATED RTD
Figure E.1: A sample calibration curve for an RTD
y = 0,9931x - 2,4944
R2 = 1
-30,00
-20,00
-10,00
0,00
10,00
20,00
30,00
40,00
50,00
-30,00 -20,00 -10,00 0,00 10,00 20,00 30,00 40,00 50,00
Referans [oC]
Ölç
üm S
onuç
ları
[ oC
]
RTD
Linear (RTD)
133
E.2 THE CALIBRATION RESULTS FOR THE CPTS
Table E.1: The calibration results for the compact pressure transducers
PSI 221 222 PSI 221 222 PSI 221 (A) 222 (A)
0 0,003983 0,003985 0 0,003982 0,003989 0 0,003983 0,0039871 0,004064 0,004063 1 0,004061 0,004065 1 0,004063 0,0040642 0,004137 0,004136 2 0,00414 0,004143 2 0,004138 0,004143 0,004212 0,004214 3 0,004216 0,004217 3 0,004214 0,0042154 0,004291 0,004295 4 0,004298 0,004299 4 0,004295 0,0042975 0,004366 0,00437 5 0,004376 0,004377 5 0,004371 0,0043736 0,004445 0,004448 6 0,004453 0,004454 6 0,004449 0,0044517 0,004524 0,004526 7 0,00453 0,00453 7 0,004527 0,0045288 0,004601 0,004602 8 0,004606 0,004607 8 0,004603 0,0046049 0,00468 0,00468 9 0,004683 0,004684 9 0,004681 0,00468210 0,004759 0,004758 10 0,004764 0,004761 10 0,004761 0,0047611 0,004838 0,004837 11 0,004841 0,004842 11 0,004839 0,0048412 0,004917 0,004915 12 0,004918 0,004917 12 0,004917 0,00491613 0,004997 0,004994 13 0,004996 0,004995 13 0,004997 0,00499414 0,005078 0,005073 14 0,005078 0,005076 14 0,005078 0,00507415 0,005152 0,005147 15 0,005156 0,005152 15 0,005154 0,0051520 0,005543 0,005536 20 0,005551 0,005545 20 0,005547 0,0055440 0,007121 0,007114 40 0,007133 0,007124 40 0,007127 0,00711960 0,008742 0,008733 60 0,008706 0,008695 60 0,008724 0,00871480 0,010273 0,01026 80 0,010282 0,010268 80 0,010277 0,010264100 0,011859 0,011846 100 0,011862 0,011847 100 0,011861 0,011846120 0,013432 0,013416 120 0,013435 0,013417 120 0,013433 0,013417140 0,01502 0,015001 140 0,015017 0,014993 140 0,015018 0,014997160 0,016588 0,016564 160 0,016597 0,016574 160 0,016592 0,016569180 0,018162 0,018146 180 0,018169 0,018148 180 0,018165 0,018147200 0,019752 0,019717 200 0,019747 0,019723 200 0,019749 0,01972
0-200 psi çıkış 0-200 psi iniş 0-200 psi ortalama
134
E.3 THE CALIBRATION CURVE FOR THE CPT 1
Figure E.2: The calibration curve for the compact pressure transducer 1
E.4: THE CALIBRATION CURVE FOR THE CPT 2
Figure E.3: The calibration curve for the compact pressure transducer 2
221
y = 12682x - 50,417
R2 = 1
0
50
100
150
200
250
0 0,005 0,01 0,015 0,02 0,025
Akım (A)
Bas
ınç
(psi
)
221 (A) Linear (221 (A))
222
y = 12705x - 50,52
R2 = 1
0
50
100
150
200
250
0 0,005 0,01 0,015 0,02 0,025
Akım (A)
Bas
ınç
(psi
)
222 (A) Linear (222 (A))
135
E.5 THE CALIBRATION RESULTS FOR THE DPT
Table E.2: The calibration results for the differential pressure transducer
E.6 THE CALIBRATION CURVE FOR THE DPT
Figure E.4: The calibration curve for the differential pressure transducer
bar A0 0,004069
0,02 0,005030,04 0,0059660,06 0,0068710,08 0,0078180,1 0,0087940,12 0,0096780,14 0,010610,16 0,0115690,18 0,0125140,2 0,0134230,22 0,0143180,24 0,0152710,26 0,0162190,28 0,0171320,3 0,0180520,32 0,0190250,34 0,019949
y = 21,452x - 0,0878
R2 = 1
-0,05
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0 0,005 0,01 0,015 0,02 0,025
Akım (A)
Bas
ınç
(bar
)
121 Linear (121)
136
APPENDIX F
WATER-ETHYLENE GLYCOL 1:1 VOLUMETRIC MIXTURE PROPERTIES
F.1: SPECIFIC GRAVITY OF THE MIXTURE
Table F.1: Specific gravity dependence on temperature for the water – ethylene
glycol mixture
Temperature Specific Gravity (°C) - -17.8 1.1 4.4 1.088
26.7 1.077 48.9 1.064
Figure F.1: Specific gravity versus temperature for the water – ethylene glycol
mixture
y = -0,0005x + 1,0906R² = 0,999
1,06
1,07
1,08
1,09
1,1
-20 0 20 40
Spe
cific
Gra
vity
Temperature [°C]
Specific Gravity
137
F.2 SPECIFIC HEAT OF THE MIXTURE
Table F.2: Specific heat dependence on temperature for the water – ethylene
glycol mixture
Temperature Specific Heat (°C) (J/kgK) -17.8 3265.704 4.4 3328.506
26.7 3412.242 48.9 3483.4176
Figure F.2: Specific heat versus temperature for the water – ethylene glycol
mixture
y = 3,3134x + 3320,9R² = 0,9974
3250
3300
3350
3400
3450
3500
-20 -10 0 10 20 30 40 50
Spe
cific
Hea
t [J/
kgK
]
Temperature [°C]
Specific Heat
138
F.3 DYNAMIC VISCOSITY OF THE MIXTURE
Table F.3: Dynamic viscosity dependence on temperature for the water –
ethylene glycol mixture
Temperature Dynamic Viscosity (°C) (kg/ms) -17.8 22 4.4 6.5
26.7 2.8 48.9 1.5
Figure F.3: Dynamic viscosity versus temperature for the water – ethylene glycol
mixture
y = 0,012x2 - 0,5379x + 8,6353R² = 1
0
6
12
18
24
-20 -10 0 10 20 30 40 50
Dyn
amic
vis
cosi
ty [
kg/m
s]
Temperature [°C]
Dynamic Viscosity
139
F.4 CONDUCTIVITY OF THE MIXTURE
Table F.4: Conductivity dependence on temperature for the water – ethylene
glycol mixture
Temperature Conductivity (°C) (W/mK) -35 0.41 40 0.4 70 0.395
Figure F.4: Conductivity versus temperature for the water – ethylene glycol
mixture
y = -0,0001x + 0,4052R² = 0,9973
0,39
0,4
0,41
-35 0 35 70
Con
duct
ivity
[W
/mK
]
Temperature [°C]
Conductivity
140
APPENDIX G
THE REFRIGERANT (R-134A) PROPERTIES
G.1 VISCOSITY (LIQUID) OF THE REFRIGERANT
Table G.1: Viscosity (liquid) dependence on saturation pressure for R-134a
Saturation pressure Viscosity (liquid) (bar) (Pa.s) 0.728 0.000425 1.159 0.00037 1.765 0.000325 2.607 0.000288 3.721 0.000256 5.175 0.00023 7.02 0.000208 9.33 0.000189 12.16 0.000172 15.59 0.000158 19.71 0.000145
141
Figure G.1: Viscosity (liquid) versus saturation pressure for R-134a
G.2 VISCOSITY (VAPOR) OF THE REFRIGERANT
Table G.2: Viscosity (vapor) dependence on saturation pressure for R-134a
Saturation pressure Viscosity (gas)
(bar) (Pa.s)
0.728 0.0000095
1.159 0.0000099
1.765 0.0000104
2.607 0.0000108
3.721 0.0000112
5.175 0.0000117
7.02 0.0000121
9.33 0.0000125
12.16 0.0000129
15.59 0.0000133
19.71 0.0000137
y = 0,0004x-0,326
R² = 0,9992
0
0,0001
0,0002
0,0003
0,0004
0,0005
0 5 10 15 20
Vis
cosi
ty (
liqu
id)
[Pa.
s]
Saturation Pressure [bar]
Viscosity (liquid)
142
Figure G.2: Viscosity (vapor) versus saturation pressure for R-134a
G.3 CONDUCTIVITY (LIQUID) OF THE REFRIGERANT
Table G.3: Conductivity (liquid) dependence on saturation pressure for R-134a
Saturation pressure Conductivity (liquid)
(bar) (W(mK)
0.728 0.099
1.159 0.095
1.765 0.091
2.607 0.087
3.721 0.083
5.175 0.079
7.02 0.075
9.33 0.071
12.16 0.068
15.59 0.064
19.71 0.06
y = 1E-05x0,1119
R² = 0,9983
6E-06
8E-06
1E-05
1,2E-05
1,4E-05
0 5 10 15 20
Vis
cosi
ty (
gas
) [P
a.s]
Saturation Pressure [bar]
Viscosity (vapor)
143
Figure G.3: Conductivity (liquid) versus saturation pressure for R-134a
G.4 CONDUCTIVITY (VAPOR) OF THE REFRIGERANT
Table G.4: Conductivity (vapor) dependence on saturation pressure for R-134a
Saturation pressure Conductivity (gas)
(bar) (W/mK)
0.728 0.008
1.159 0.008
1.765 0.008
2.607 0.009
3.721 0.009
5.175 0.01
7.02 0.01
9.33 0.01
12.16 0.011
15.59 0.011
19.71 0.012
y = -0,012ln(x) + 0,0972
R² = 0,9916
0,05
0,06
0,07
0,08
0,09
0,1
0 5 10 15 20
Co
nd
uct
ivit
y (l
iqu
id)
[W/m
K]
Saturation Pressure [bar]
Conductivity (liquid)
144
Figure G.4: Conductivity (vapor) versus saturation pressure for R-134a
G.5 PRANDTL NUMBER (LIQUID) OF THE REFRIGERANT
Table G.5: Pr (liquid) dependence on saturation pressure for R-134a
Saturation pressure Pr (liquid)
(bar) -
0.728 4.99
1.159 4.72
1.765 4.49
2.607 4.31
3.721 4.17
5.175 4.07
7.02 4.00
9.33 3.98
y = 0,0002x + 0,0082
R² = 0,9091
0,006
0,008
0,01
0,012
0,014
0 5 10 15 20
Co
nd
uct
ivit
y (g
as)
[W/m
K]
Saturation Pressure [bar]
Conductivity (vapor)
145
Figure G.5: Pr (liquid) versus saturation pressure for R-134a
G.6 PRANDTL NUMBER (VAPOR) OF THE REFRIGERANT
Table G.6: Pr (vapor) dependence on saturation pressure for R-134a
Saturation pressure Pr (gas)
(bar) -
0.728 0.9
1.159 0.96
1.765 1.02
2.607 1.08
3.721 1.14
5.175 1.2
7.02 1.27
9.33 1.34
12.16 1.57
15.59 1.44
19.71 1.74
y = 4,7684x-0,091
R² = 0,9717
3,8
4
4,2
4,4
4,6
4,8
5
0 2 4 6 8 10
Pra
nd
tl(l
iqu
id)
Saturation Pressure [bar]
Prandtl (liquid)
146
Figure G.6: Pr (vapor) versus saturation pressure for R-134a
G.7 SURFACE TENSION OF THE REFRIGERANT
Table G.7: Surface tension dependence on saturation pressure for R-134a
Saturation pressure Surface tension
(bar) (N/m)
1.159 0.0145
1.765 0.0131
2.607 0.0117
3.721 0.0103
5.175 0.009
y = 0,9373x0,1546
R² = 0,9971
0,8
0,9
1
1,1
1,2
1,3
1,4
0 2 4 6 8 10
Pra
nd
tl(g
as)
Saturation Pressure [bar]
Prandtl (vapor)
147
Figure G.7: Surface tension versus saturation pressure for R-134a
G.8 SPECIFIC HEAT (LIQUID) OF THE REFRIGERANT
Table G.8: Specific heat (liquid) dependence on saturation pressure for R-134a
Saturation pressure Specific heat (liquid)
(bar) (J/kgK)
0.728 1162
1.159 1212
1.765 1259
2.607 1306
3.721 1351
5.175 1397
7.02 1446
9.33 1497
12.16 1559
15.59 1638
19.71 1750
y = 0,0155x-0,318
R² = 0,9887
0,008
0,01
0,012
0,014
0,016
0 2 4 6 8 10
Su
rfac
e T
ensi
on
[N
/m]
Saturation Pressure [bar]
Surface Tension
148
Figure G.8: Specific heat (liquid) versus saturation pressure for R-134a
y = 0,1296x3 - 4,5893x2 + 71,613x + 1133,7
R² = 0,9963
1000
1200
1400
1600
1800
0 5 10 15 20
Sp
ecif
ic h
eat
(liq
uid
) [J
/kg
K]
Saturation Pressure [bar]
Specific heat(liquid)
149
APPENDIX H
THE RESULTS FOR A SAMPLE TWO-PHASE EXPERIMENT
H.1 THE GEOMETRY OF BOTH SIDES
Table H.1: The geometries of both sides
Refrigerant side geometry Water side geometry
Inner diameter mm 1.65 Inner diameter mm 3.175
Outer diameter mm 3.175 Outer diameter mm 10
Inner perimeter mm 5.18 Inner perimeter mm 9.97
Cross sectional area mm² 2.14 Outer perimeter mm 31.42
Test length mm 400 Cross sectional area mm² 70.62
Heat transfer area mm² 2073.45 Hydraulic diameter mm 6.83
Equivalent diameter mm 28.32
150
H.2 TEST CONDITIONS FOR THE EXPERIMENT
Table H.2: Test conditions for the experiment
Refrigerant mass flow rate g/s 0.505
Refrigerant inlet pressure bar 7.32
Refrigerant flowmeter °C 16.96
Refrigerant inlet temperature °C 28.61
Refrigerant exit temperature °C 28.09
Refrigerant Pressure drop bar 0.05
Water inlet temperature °C 33.83
Water exit temperature °C 33.70
Water volumetric flow rate l/min 1.65
Water mass flow rate kg/s 0.030
Water mass flow rate g/s 29.53
Test inlet wall temperature °C 30.55
Test exit wall temperature °C 30.71
Pre-heater power input W 28.30
Pre-heater net power W 24.05
Viscous Heating W 9.06
Net heat transfer rate W 22.24
151
H.3 REFRIGERANT EXPERIMENT STATES
Table H.3: Refrigerant experiment states
Flowmeter State (Pre-heater inlet)
Refrigerant pressure bar 7.32
Refrigerant temperature °C 16.96
Refrigerant quality (%) - -
Refrigerant enthalpy (kJ/kg) kJ/kg 223.23
Test Inlet (Pre-heater exit)
Refrigerant pressure bar 7.32
Refrigerant temperature °C 28.24
Refrigerant quality (%) - 18.09
Refrigerant enthalpy (kJ/kg) kJ/kg 270.81
Test Exit
Refrigerant pressure bar 7.27
Refrigerant temperature °C 28.02
Refrigerant quality (%) - 43.40
Refrigerant enthalpy (kJ/kg) kJ/kg 314.81
H.4 WATER SIDE PROPERTIES
Table H.4: Water side properties
Velocity m/s 0.3894
Density kg/m³ 1073.72
Viscosity Pa.s 0.0042
Reynolds - 687
Specific heat J/kgK 3432.78
Conductivity W/mK 0.402
Pr - 35.49
Average temperature °C 33.77
Mass flow rate kg/s 0.0295
152
H.5 REFRIGERANT EXPERIMENTAL CALCULATIONS
Table H.5: Refrigerant experimental calculations
Net heat transfer rate W 22.24
Heat flux W/m² 10726.16
Mass flux kg/m²s 236.40
Water average temperature °C 33.77
Wall average temperature °C 30.63
Water - wall temp. difference °C 3.14
Refrigerant inlet temperature °C 28.24
Refrigerant exit temperature °C 28.02
Temp. Diff. Inlet °C -2.39
Temp. Diff. Exit °C -2.61
Log. Mean Temp. Diff. °C -2.50
U W/m²K 4293.24
h W/m²K 4318.26
Refrigerant average temp. °C 28.13
Refrigerant saturation pres. bar 7.30
Average quality (%) - 30.75
153
H.6 CALCULATIONS OF THE CORRELATIONS IN THE LITERATURE
Table H.6: Calculations of correlations in the literature
CHEN
ρliquid kg/m³ 1230.78 Mass flow rate kg/s 5.05E-04
ρvapor kg/m³ 26.08 µ Pa.s 1.52E-04
µliquid Pa.s 2.091E-04 Refrigerant Re - 2624
µvapor Pa.s 1.249E-05 Reliquid - 1865
Xtt - 0.45 Retwo-phase - 8511
1/Xtt - 2.22 Peliquid - 7420.77
F - 4.52 Peliquid*d/L - 30.61
Fo - 3.13 hliquid only W/m²K 163.59
kliquid W/mK 0.0737 hconvective boiling W/m²K 511.89
kvapor W/mK 0.0097 θB K 2.50
Prliquid - 3.98 ∆Pv J/m³ 39809.22
Prvapor - 1.27 S - 0.91
σ N/m 0.0077 hvapor only W/m²K 2398.88
cp,liquid J/kgK 1462.25 h W/m²K 2910.77
GÜNGÖR BERTSCH
Reliquid - 1292 hnucleate boiling W/m²K 4088.31
E - 5.28 Reliquid - 1865
hliquid W/m²K 163.59 Revapor - 31228
S - 0.88 hconvection, liquid W/m²K 251.54
hp W/m²K 4088.34 hconvection, vapor W/m²K 78.75
hTP W/m²K 4449.99 hconvection, boiling W/m²K 198.41
Co - 0.491
h W/m²K 4137.56
154
H.7 EXPERIMENTAL RESULTS AND RESULTS OF THE CORRELATIONS IN
LITERATURE
Table H.7: Experimental results and the results of the correlations in literature
Refrigerant mass flow rate g/s 0.51
Mass flux kg/m²s 236.40
Average quality (%) - 30.75
hexp W/m²K 4318.26
hGüngör W/m²K 4449.99
hBertsch W/m²K 4137.56
hChen W/m²K 2910.769
Refrigerant pressure drop bar 0.05